Professional Documents
Culture Documents
CH 19
CH 19
CH 19
Table of Contents
19.1 Introduction
This chapter provides information intended for prestressed I-girders. Prestressed box girders
and general prestressed concrete guidelines are also included in this chapter.
The definition of prestressed concrete as given by the ACI Committee on Prestressed Concrete
is:
"Concrete in which there has been introduced internal stresses of such magnitude and
distribution that the stresses resulting from given external loadings are counteracted to
a desired degree. In reinforced concrete members the prestress is commonly
introduced by tensioning the steel reinforcement.”
This internal stress is induced into the member by either of the following prestressing methods.
19.1.1 Pretensioning
In pretensioning, the tendons are first stressed to a given level and then the concrete is cast
around them. The tendons may be composed of wires, bars or strands.
The most common system of pretensioning is the long line system, by which a number of units
are produced at once. First the tendons are stretched between anchorage blocks at opposite
ends of the long stretching bed. Next the spacers or separators are placed at the desired
member intervals, and then the concrete is placed within these intervals. When the concrete
has attained a sufficient strength, the steel is released and its stress is transferred to the
concrete via bond.
19.1.2 Post-Tensioning
In post-tensioning, the concrete member is first cast with one or more post-tensioning ducts or
tubes for future insertion of tendons. Once the concrete is sufficiently strong, the tendons are
stressed by jacking against the concrete. When the desired prestress level is reached, the
tendons are locked under stress by means of end anchorages or clamps. Subsequently, the
duct is filled with grout to protect the steel from corrosion and give the added safeguard of
bond.
This section defines the internal stress that results from either prestressing method.
Figure 19.2-1
Simple Span Prestressed Concrete Beam
The horizontal component, P, of the tendon force, F, is assumed constant at any section along
the length of the beam.
Also, at any section of the beam the forces in the beam and in the tendon are in equilibrium.
Forces and moments may be equated at any section.
Figure 19.2-2
Assumed Sign Convention for Section Forces
The assumed sign convention is as shown in Figure 19.2-2 with the origin at the intersection
of the section plane and the center of gravity (centroidal axis) of the beam. This convention
indicates compression as positive and tension as negative.
The eccentricity of the tendon can be either positive or negative with respect to the center of
gravity; therefore it is unsigned in the general equation. The reaction of the tendon on the beam
is always negative; therefore the horizontal component is signed as:
P = F cos θ
Then, by equating forces in the x-direction, the reaction, P, of the tendon on the concrete
produces a compressive stress equal to:
P
f1 =
A
Where:
Since the line of action of the reaction, P, is eccentric to the centroidal axis of the beam by the
amount e, it produces a bending moment.
M = Pe
This moment induces stresses in the beam given by the flexure formula:
My Pey
f2 = =
I I
Where:
y = Distance from the centroidal axis to the fiber under consideration, with
an unsigned value in the general equations
I = Moment of inertia of the section about its centroidal axis
The algebraic sum of f1 and f2 yields an expression for the total prestress on the section when
the beam is not loaded.
P Pey
f p = f1 + f 2 = +
A I
Now, by substituting I = Ar2, where r is the radius of gyration, into the above expression and
arranging terms, we have:
P ey
fp = 1 + 2
A r
Figure 19.2-3
Calculation of Concrete Stress Due to Prestress Force
Finally, we equate forces in the y-direction which yields a shear force, V, over the section of
the beam due to the component of the tendon reaction.
V = F sin θ = P tan θ
This section outlines several important considerations associated with the design of
conventional pretensioned members.
The actual required compressive strength of the concrete at prestress transfer, f’ci, is to be
stated on the plans. For typical prestressed girders, f’ci(min) is 0.75(f’c).
For prestressed I-girders, the use of concrete with strength greater than 8 ksi is only allowed with
the prior approval of the BOS Development Section. Occasional use of strengths up to 8.5 ksi
may be allowed. Strengths exceeding these values are difficult for local fabricators to consistently
achieve as the coarse aggregate strength becomes the controlling factor.
For prestressed box girders, the use of concrete with strength greater than 5 ksi is only allowed
with prior approval of the BOS Development Section.
The use of 8 ksi concrete for prestressed I-girders and 6.8 ksi for f’ci still allows the fabricator
to use a 24-hour cycle for girder fabrication. There are situations in which higher strength
concrete in the prestressed I-girders may be considered for economy, provided that f’ci does
not exceed 6.8 ksi. Higher strength concrete may be considered if the extra strength is needed
to avoid using a less economical superstructure type or if a shallower girder can be provided
and its use justified for sufficient reasons (min. vert. clearance, etc.) Using higher strength
concrete to eliminate a girder line is not the preference of the Bureau of Structures. It is often
more economical to add an extra girder line than to use debonded strands with the minimum
number of girder lines. After the number of girders has been determined, adjustments in girder
spacing should be investigated to see if slab thickness can be minimized and balance between
interior and exterior girders optimized.
Prestressed I-girders below the required 28-day concrete strength (or 56-day concrete strength
for f’c = 8 ksi) will be accepted if they provide strength greater than required by the design and
at the reduction in pay schedule in the Wisconsin Standard Specifications for Highway and
Structure Construction.
January 2017 19-7
WisDOT Bridge Manual Chapter 19 – Prestressed Concrete
The loads that a member is subjected to during its design life and those stages that generally
influence the design are discussed in LRFD [5.9] and in the following sections. The allowable
stresses at different loading stages are defined in LRFD [5.9.3] and LRFD [5.9.4].
Prestress transfer is the initial condition of prestress that exists immediately following the
release of the tendons (transfer of the tendon force to the concrete). The eccentricity of the
prestress force produces an upward camber. In addition, a stress due to the dead load of the
member itself is also induced. This is a stage of temporary stress that includes a reduction in
prestress due to elastic shortening of the member.
19.3.2.2 Losses
After elastic shortening losses, the external loading is the same as at prestress transfer.
However, the internal stress due to the prestressing force is further reduced by losses resulting
from relaxation due to creep of the prestressing steel together with creep and shrinkage of the
concrete. It is assumed that all losses occur prior to application of service loading.
LRFD [5.9.5] provides guidance about prestress losses for both pretensioned and post-
tensioned members. This section presents a refined and approximate method for the
calculation of time-dependent prestress losses such as concrete creep and shrinkage and
prestressing steel relaxation.
WisDOT policy is to use the approximate method described in LRFD [5.9.5.3] to determine time-
dependent losses, since this method does not require the designer to assume the age of the
concrete at the different loading stages.
Losses for pretensioned members that are considered during design are listed in the following
sections.
Per LRFD [5.9.5.2.3a], the loss due to elastic shortening, ∆f pES1 (ksi), in pretensioned concrete
members shall be taken as:
Ep
∆f pES1= fcgp
Ect
Where:
Per LRFD [5.9.5.3], an estimate of the long-term losses due to steel relaxation as well as
concrete creep and shrinkage on standard precast, pretensioned members shall be taken as:
fpi A ps
∆f pLT = 10.0 γ h γ st + 12.0 γ h γ st + ∆fpR
Ag
Where:
γ h = 1.7 − 0.01H
5
γ st =
(1 + f ' ci )
The losses due to elastic shortening must then be added to these time-dependent losses to
determine the total losses. For non-standard members with unusual dimensions or built using
staged segmental construction, the refined method of LRFD [5.9.5.4] shall be used. For
prestressed box girders time-dependent losses shall be determined using the refined method
of LRFD [5.9.5.4].
Fabrication losses are not considered by the designer, but they affect the design criteria used
during design. Anchorage losses which occur during stressing and seating of the prestressed
strands vary between 1% and 4%. Losses due to temperature change in the strands during
cold weather prestressing are 6% for a 60°F change. The construction specifications permit a
5% difference in the jack pressure and elongation measurement without any adjustment.
During service load, the member is subjected to the same loads that are present after prestress
transfer and losses occur, in addition to the effects of the prestressed I-girder and prestressed
box girder load-carrying behavior described in the next two sections.
In the case of a prestressed I-girder, the dead load of the deck and diaphragms are always
carried by the basic girder section on a simple span. At strand release, the girder dead load
moments are calculated based on the full girder length. For all other loading stages, the girder
dead load moments are based on the span length. This is due to the type of construction used
(that is, unshored girders simply spanning from one substructure unit to another for single-
span as well as multi-span structures).
The live load plus dynamic load allowance along with any superimposed dead load (curb,
parapet or median strip which is placed after the deck concrete has hardened) are carried by
the continuous composite section.
The standard pier diaphragm is considered to satisfy the requirements of LRFD [5.14.1.4.5] and
shall be considered to be fully effective.
In the case of multi-span structures with fully effective diaphragms, the longitudinal distribution
of the live load, dynamic load allowance and superimposed dead loads are based on a
continuous span structure. This continuity is achieved by:
b. Casting concrete between and around the abutting ends of adjacent girders to
form a diaphragm at the support. Girders shall be in line at interior supports and
equal numbers of girders shall be used in adjacent spans. The use of variable
numbers of girders between spans requires prior approval by BOS.
If the span length ratio of two adjacent spans exceeds 1.5, the girders are designed as simple
spans. In either case, the stirrup spacing is detailed the same as for continuous spans and bar
steel is placed over the supports equivalent to continuous span design. It should be noted that
this value of 1.5 is not an absolute structural limit.
In the case of prestressed box girders with a thin concrete overlay, the dead load together with
the live load and dynamic load allowance are carried by the basic girder section.
When this girder type has a composite section, the dead load of the deck is carried by the
basic section and the live load, dynamic load allowance and any superimposed dead loads are
carried by the composite section. A composite section shall consist of a reinforced deck, 6”
minimum thickness, with composite shear reinforcement extending into the deck.
The use of prestressed box girders is subject to prior-approval by the Bureau of Structures. These
structures are currently limited to the following requirements:
● Single spans
At the final stage, the factored flexural resistance of the composite section is considered. Since
the member is designed on a service load basis, it must be checked for its factored flexural
resistance at the Strength I limit state. See section 17.2.3 for a discussion on limit states.
The need for both service load and strength computations lies with the radical change in a
member's behavior when cracks form. Prior to cracking, the gross area of the member is
effective. As a crack develops, all the tension in the concrete is picked up by the reinforcement.
If the percentage of reinforcement is small, there is very little added capacity between cracking
and failure.
At the final stage, the member is checked for the Fatigue I limit state. See section 17.2.3 for a
discussion on limit states. Allowable compressive stresses in the concrete and tensile stresses
in the non-prestressed reinforcement are checked.
The intent of this section is to provide the designer with a general outline of steps for the design
of pretensioned members. Sections of interest during design include, but are not limited to, the
following locations:
• 10th points
• Hold-down points
• Regions where the prestress force changes (consider the effects of transfer and
development lengths, as well as the effects of debonded strands)
The designer must consider the amount of prestress force at each design section, taking into
account the transfer length and development length, if appropriate.
A trial prestressed I-girder arrangement is made by using Table 19.3-1 and Table 19.3-2 as a
guide. An ideal spacing results in equal strands for interior and exterior girders, together with
an optimum slab thickness. Current practice is to use a minimum haunch of (1-1/4” plus deck
cross slope times one-half top flange width) for section property calculations and then use a 3”
average haunch for concrete preliminary quantity calculations. After preliminary design this
value should be revised as needed as outlined in 19.3.4. The maximum slab overhang
dimensions are detailed in 17.6.2.
For prestressed I-girder bridges, other than pedestrian or other unusual structures, four or
more girders shall be used.
The prestressed box girder is used in an adjacent multi-beam system only. Precast units are
placed side by side and locked (post-tensioned) together. The span length, desired roadway
width and live loading control the size of the member.
When selecting a 3' wide section vs. 4' wide section, do not mix 3’ wide and 4’ wide sections
across the width of the bridge. Examine the roadway width produced by using all 3’ wide
sections or all 4’ wide sections and choose the system that is the closest to but greater than
the required roadway width. While 3’ wide sections may produce a slightly narrower roadway
width 4’ wide sections are still preferred since they require fewer sections. Verify the required
roadway width is possible when considerations are made for the roadway cross-slope. Table
19.3-3 states the approximate span limitations for each section depth. Coordinate roadway
width with roadway designers and consider some variability. See the Standards for
prestressed box girder details.
The dead load moments and shears due to the girder and concrete deck are computed for
simple spans. When superimposed dead loads are considered, the superimposed dead load
moments are based on continuous spans.
A superimposed dead load of 20 psf is to be included in all designs which account for a possible
future concrete overlay wearing surface. The future wearing surface shall be applied between
the faces of curbs or parapets and shall be equally distributed among all the girders in the
cross section.
For a cross section without a sidewalk, any curb or parapet dead load is distributed equally to
all girders.
For a cross section with a sidewalk and barrier on the overhang, sidewalk and barrier dead
loads shall be applied to the exterior girder by the lever rule. These loads shall also be applied
to the interior girder by dividing the weight equally among all the girders. A more detailed
discussion of dead load distribution can be found in 17.2.8.
The HL-93 live load shall be used for all new bridges. Refer to section 17.2.4.2 for a detailed
description of the HL-93 live load, including the design truck, design tandem, design lane, and
double truck.
The live load distribution factors shall be computed as specified in LRFD [4.6.2.2]. Table 17.2-
7 summarizes the equations required for prestressed I-girders. The moment and shear
distribution factors for prestressed I-girders are determined using equations that consider
girder spacing, span length, deck thickness, the number of girders, skew and the longitudinal
stiffness parameter. See the WisDOT policy item for live load distribution factors for
prestressed box girders.
Separate shear and moment distribution factors are computed for interior and exterior girders.
The applicability ranges of the distribution factors shall also be considered. If the applicability
ranges are not satisfied, then conservative assumptions must be made based on sound
engineering judgment.
The typical cross section for prestressed box girders shall be type “g” as illustrated in LRFD [Table
4.6.2.2.1-1].
For prestressed box girders, the St. Venant torsional inertia, J, may be calculated as closed thin-
walled sections for sections with voids, and as solid sections for sections without voids in
accordance with LRFD [C4.6.2.2.1].
The dynamic load allowance, IM, is given by LRFD [3.6.2]. Dynamic load allowance equals
33% for all live load limit states except the fatigue limit state and is not applied to pedestrian
loads or the lane load portion of the HL-93 live load. See 17.2.4.3 for further information
regarding dynamic load allowance.
The design of concrete decks on prestressed I-girders is based on LRFD [4.6.2.1]. Moments
from truck wheel loads are distributed over a width of deck which spans perpendicular to the
girders. This width is known as the distribution width and is given by LRFD [Table 4.6.2.1.3-
1]. See 17.5 for further information regarding deck design.
The effective flange width is the width of the deck slab that is to be taken as effective in
composite action for determining resistance for all limit states. The effective flange width, in
accordance with LRFD [4.6.2.6], is equal to the tributary width of the girder for interior girders.
For exterior girders, it is equal to one half the effective flange width of the adjacent interior
girder plus the overhang width. The effective flange width shall be determined for both interior
and exterior beams.
Since the deck concrete has a lower strength than the girder concrete, it also has a lower
modulus of elasticity. Therefore, when computing composite section properties, the effective
flange width (as stated above) must be reduced by the ratio of the modulus of elasticity of the
deck concrete divided by the modulus of elasticity of the girder concrete.
WisDOT uses the formulas shown below to determine Ec for prestressed girder design. For 6 ksi
girder concrete, Ec is 5,500 ksi, and for 4 ksi deck concrete, Ec is 4,125 ksi. The Ec value of 5,500
ksi for 6 ksi girder concrete strength was determined from deflection studies. These equations are
used in place of those presented in LRFD [5.4.2.4] for the following calculations: strength, section
properties, and deflections due to externally applied dead and live loads.
For slab concrete strength other than 4 ksi, Ec is calculated from the following formula:
4,125 f ' c
Ec = (ksi)
4
For girder concrete strengths other than 6 ksi, Ec is calculated from the following formula:
5,500 f 'c
Ec = (ksi)
6
WisDOT uses the equation presented in LRFD [5.4.2.4] (and shown below) to calculate the
modulus of elasticity at the time of release using the specified value of f’ci. This value of Ei is
used for loss calculations and for girder camber due to prestress forces and girder self weight.
Ec = 33,000 ⋅ K1 ⋅ w c
1. 5
f 'ci
January 2017 19-14
WisDOT Bridge Manual Chapter 19 – Prestressed Concrete
Where:
In many cases, stress at the Service III limit state in the bottom fiber at or near midspan after
losses will control the flexural design. Determine a trial strand pattern for this condition and
proceed with the flexural design, adjusting the strand pattern if necessary.
The design stress is the sum of the Service III limit state bottom fiber stresses due to non-
composite dead load on the basic girder section, plus live load, dynamic load allowance and
superimposed dead load on the composite section, as follows:
Where:
The point of maximum stress is generally 0.5 of the span for both end and intermediate spans.
But for longer spans (over 100'), the 0.4 point of the end span may control and should be
checked.
With fdes known, compute the required effective stress in the prestressing steel after losses, fpe,
needed to counteract all the design stress except an amount of tension equal to the tensile
stress limit listed in LRFD [Table 5.9.4.2.2-1]. The top of the girder is subjected to severe
corrosion conditions and the bottom of the girder is subjected to moderate exposure. The
Service III tensile stress at the bottom fiber after losses for pretensioned concrete shall not
exceed 0.19λ f ' c (or 0.6 ksi) ; where λ = concrete density modification factor LRFD [5.4.2.8],
and has a value of 1.0 for normal weight concrete. Therefore:
Note: A conservative approach used in hand calculations is to assume that the allowable
tensile stress equals zero.
Ppe ey
fpe = 1 + 2
A r
Where:
r = I
of the basic beam (in)
A
For prestressed box girders, assume an e and apply this to the above equation to determine
Ppe and the approximate number of strands. Then a trial strand pattern is established using the
Standard Details as a guide, and a check is made on the assumed eccentricity. For prestressed
I-girders, fpe is solved for several predetermined patterns and is tabulated in the Standard
Details.
Present practice is to detail all spans of equal length with the same number of strands, unless
a span requires more than three additional strands. In this case, the different strand
arrangements are detailed along with a plan note stating: "The manufacturer may furnish all
girders with the greater number of strands."
Several checks need to be performed at the service limit state. Refer to the previous narrative
in 19.3.3 for sections to be investigated and section 17.2.3.2 for discussion on the service limit
state. Note that Service I limit state is used when checking compressive stresses and Service
III limit state is used when checking tensile stresses.
• Verify that the Service III tensile stress due to beam self-weight and prestress applied
to the basic beam at transfer does not exceed the limits presented in LRFD [Table
5.9.4.1.2-1], which depend upon whether or not the strands are bonded and satisfy
stress requirements. This will generally control at the top of the beam near the beam
ends where the dead load moment approaches zero and is not able to counter the
tensile stress at the top of the beam induced by the prestress force. When the
calculated tensile stress exceeds the stress limits, the strand pattern must be modified
by draping or partially debonding the strand configuration.
• Verify that the Service I compressive stress due to beam self-weight and prestress
applied to the basic beam at transfer does not exceed 0.65 f’ci, as presented in LRFD
[5.9.4.1.1]. This will generally control at the bottom of the beam near the beam ends or
at the hold-down point if using draped strands.
• Verify that the Service III tensile stress due to all dead and live loads applied to the
appropriate sections after losses does not exceed the limits presented in LRFD [Table
5.9.4.2.2-1]. No tensile stress shall be permitted for unbonded strands. The tensile
stress of bonded strands shall not exceed 0.19λ f ' c (or 0.6 ksi) as all strands shall
be considered to be in moderate corrosive conditions. This will generally control at the
bottom of the beam near midspan and at the top of the continuous end of the beam.
The value of λ is 1.0 for normal weight concrete LRFD [5.4.2.8].
• Verify that the Service I compressive stress due to all dead and live loads applied to
the appropriate sections after losses does not exceed the limits presented in LRFD
[Table 5.9.4.2.1-1]. Two checks need to be made for girder bridges. The compressive
stress due to the sum of effective prestress and permanent loads shall not exceed 0.45
f’c (ksi). The compressive stress due to the sum of effective prestress, permanent loads
and transient loads shall not exceed 0.60φ w f ' c (ksi). The term φ w , a reduction factor
applied to thin-walled box girders, shall be 1.0 for WisDOT standard girders.
• Verify that Fatigue I compressive stress due to fatigue live load and one-half the sum
of effective prestress and permanent loads does not exceed 0.40 f’c (ksi) LRFD
[5.5.3.1].
• Verify that the Service I compressive stress at the top of the deck due to all dead and
live loads applied to the appropriate sections after losses does not exceed 0.40 f’c.
The top of the prestressed I-girders at interior supports shall be designed as reinforced concrete
members at the strength limit state in accordance with LRFD [5.14.1.4.6]. In this case, the stress
limits for the service limit state shall not apply to this region of the precast girder.
When straight strands are bonded for the full length of a prestressed girder, the tensile and
compressive stresses near the ends of the girder will likely exceed the allowable service limit
state stresses. This occurs because the strand pattern is designed for stresses at or near
midspan, where the dead load moment is highest and best able to balance the effects of the
prestress. Near the ends of the girder this dead load moment approaches zero and is less able
to balance the prestress force. This results in tensile stresses in the top of the girder and
compressive stresses in the bottom of the girder. The allowable initial tensile and compressive
stresses are presented in the first two bullet points of 19.3.3.11. These stresses are a function
of f'ci, the compressive strength of concrete at the time of prestress force transfer. Transfer and
development lengths should be considered when checking stresses near the ends of the
girder.
The designer should start with a straight (raised), fully bonded strand pattern. If this
overstresses the girder near the ends, the following methods shall be utilized to bring the girder
within the allowable stresses. These methods are listed in order of preference and discussed
in the following sections:
1. Use raised strand pattern (If excessive top flange reinforcement or if four or more
additional strands versus a draped strand pattern are required, consider the draped
strand alternative)
Only show one strand pattern per span (i.e. Do not show both raised and draped span
alternatives for a given span).
A different girder spacing may need to be selected. It is often more economical to add an extra
girder line than to maximize the number of strands and use debonding.
Prestressed box girders strands are to be straight, bonded, and located as shown in the
Standard Details.
Some of the standard strand patterns listed in the Standard Details show a raised strand
pattern. Generally strands are placed so that the center of gravity of the strand pattern is as
close as possible to the bottom of the girder. With a raised strand pattern, the center of gravity
of the strand pattern is raised slightly and is a constant distance from the bottom of the girder
for its entire length. Present practice is to show a standard raised arrangement as a preferred
alternate to draping for short spans. For longer spans, debonding at the ends of the strands is
an alternate (see 19.3.3.12.3). Use 0.6” strands for all raised patterns.
Draping some of the strands is another available method to decrease stresses from prestress
at the ends of the I-beam where the stress due to applied loads are minimum.
The typical strand profile for this technique is shown in Figure 19.3-1.
Figure 19.3-1
Typical Draped Strand Profile
Note that all the strands that lie within the “vertical web zone” of the mid-span arrangement are
used in the draped group.
The engineer should show only one strand size for the draped pattern on the plans. Use only
0.5” strands for the draped pattern on 28” and 36” prestressed I-girders and 0.6” strands for all
raised (straight) patterns for these shapes. Use 0.6” strands, only, for 36W”, 45W”, 54W”,
72W” and 82W” prestressed I-girders. See Chapter 40 standards for 45”, 54” and 70”
prestressed I-girders.
Hold-down points for draped strands are located approximately between the 1/3 point and the
4/10 point from each end of the girder. The Standard Details, Prestressed Girder Details, show
B values at the 1/4 point of the girder. On the plan sheets provide values for Bmin and Bmax as
determined by the formulas shown on the Standards.
The maximum slope specified for draped strands is 12%. This limit is determined from the safe
uplift load per strand of commercially available strand restraining devices used for hold-downs.
The minimum distance, D, allowed from center of strands to top of flange is 2”. For most
designs, the maximum allowable slope of 12% will determine the location of the draped
strands. Using a maximum slope will also have a positive effect on shear forces.
Initial girder stresses are checked at the end of the transfer length, which is located 60 strand
diameters from the girder end. The transfer length is the embedment length required to develop
fpe, the effective prestressing steel stress (ksi) after losses. The prestressing steel stress varies
linearly from 0.0 to fpe along the transfer length.
The longer full development length of the strand is required to reach the larger prestressing
steel stress at nominal resistance, fps (ksi). The strand stress is assumed to increase linearly
from fpe to fps over the distance between the transfer length and development length.
2
l d
≥ κ fps − fpe d b
3
Where:
Development length
Transfer length
Steel Stress
Linear
fps
fpe
Free 60db
end of
strand ld
Figure 19.3-2
Transfer and Development Length
The designer may use debonded strands if a raised or draped strand configuration fails to meet
the allowable service stresses. The designer should exercise caution when using debonded
strands as this may not result in the most economical design. Partially debonded strands are
fabricated by wrapping sleeves around individual strands for a specified length from the ends
of the girder, rendering the bond between the strand and the girder concrete ineffective for the
wrapped, or shielded, length.
Bond breakers should only be applied to interior strands as girder cracking has occurred when
they were applied to exterior strands. In computing bond breaker lengths, consideration is
January 2017 19-20
WisDOT Bridge Manual Chapter 19 – Prestressed Concrete
given to the theoretical stresses at the ends of the girder. These stresses are due entirely to
prestress. As a result, the designer may compute a stress reduction based on certain strands
having bond breakers. This reduction can be applied along the length of the debonded strands.
Partially debonded strands must adhere to the requirements listed in LRFD [5.11.4.3]. The list
of requirements is as follows:
• The number of debonded strands shall not exceed 25% of the total number of strands.
• The number of debonded strands in any horizontal row shall not exceed 40% of the
strands in that row.
• The length of debonding shall be such that all limit states are satisfied with
consideration of the total developed resistance (transfer and development length) at
any section being investigated.
• Not more than 40% of the debonded strands, or four strands, whichever is greater,
shall have debonding terminated at any section.
• The strand pattern shall be symmetrical about the vertical axis of the girder. The
consideration of symmetry shall include not only the strands being debonded but their
debonded length as well, with the goal of keeping the center of gravity of the prestress
force at the vertical centerline of the girder at any section. If the center of gravity of the
prestress force deviates from the vertical centerline of the girder, the girder will twist,
which is undesirable.
• Exterior strands in each horizontal row shall be fully bonded for crack control purposes.
The design factored positive moment is determined using the following equation:
The Strength I limit state is applied to both simple and continuous span structures. See 17.2.4
for further information regarding loads and load combinations.
The nominal flexural resistance assuming rectangular behavior is given by LRFD [5.7.3.2.3]
and LRFD [5.7.3.2.2].
The section will act as a rectangular section as long as the depth of the equivalent stress block,
a, is less than or equal to the depth of the compression flange (the structural deck thickness).
Per LRFD [5.7.3.2.2]:
a = cβ 1
Where:
By neglecting the area of mild compression and tension reinforcement, the equation presented
in LRFD [5.7.3.1.1] for rectangular section behavior reduces to:
A ps f pu
c=
f pu
α1f 'c β1b + kA ps
dp
Where:
α1 = Stress block factor; equals 0.85 (for f’C < 10.0 ksi) LRFD [5.7.2.2]
c
N.A.
bw
dp
Aps
Figure 19.3-3
Depth to Neutral Axis, c
Verify that rectangular section behavior is allowed by checking that the depth of the equivalent
stress block, a, is less than or equal to the structural deck thickness. If it is not, then T-section
behavior provisions should be followed. If the T-section provisions are used, the compression
block will be composed of two different materials with different compressive strengths. In this
situation, LRFD [C5.7.2.2] recommends using β 1 and α1 corresponding to the lower f’c. The
following equation for c shall be used for T-section behavior: LRFD [5.7.3.1.1]
A ps f pu − α1f 'c (b − b w )h f
c=
f
α1f 'c β1b w + kA ps pu
dp
Where:
bw = Width of web (in) – use the top flange width if the compression block
does not extend below the haunch.
hf = Depth of compression flange (in)
The factored flexural resistance presented in LRFD [5.7.3.2.2] is simplified by neglecting the
area of mild compression and tension reinforcement. Furthermore, if rectangular section
behavior is allowed, then bw = b, where bw is the web width as shown in Figure 19.3-3. The
equation then reduces to:
a
Mr = φA ps fps d p −
2
Where:
If the T-section provisions must be used, the factored moment resistance equation is then:
a a h
M r = φA ps f ps d p − + α1φf 'c (b − b w )h f − f
2 2 2
Where:
The engineer must then verify that Mr is greater than or equal to Mu.
WisDOT standard prestressed I-girders and strand patterns are tension-controlled. The ε t check,
as specified in LRFD [5.7.2.1], is not required when the standard girders and strand patterns are
used, and φ = 1.
Per LRFD [5.7.3.3.2], the minimum amount of prestressed reinforcement provided shall be
adequate to develop a Mr at least equal to the lesser of Mcr, or 1.33Mu.
Where:
Per LRFD [5.4.2.6], the modulus of rupture for normal weight concrete is given by:
The amount of reinforcement is that which is sufficient to resist the total tension force in the
concrete based on the assumption of an uncracked section.
For draped designs, the control is at the hold-down point of the girder. At the hold-down point,
the initial prestress is acting together with the girder dead load stress. This is where tension
due to prestress is still maximum and compression due to girder dead load is decreasing.
For non-draped designs, the control is at the end of the member where prestress tension exists
but dead load stress does not.
Note that a minimum amount of reinforcement is specified in the Standards. This is intended
to help prevent serious damage due to unforeseeable causes like improper handling or storing.
The horizontal shear reinforcement resists the Strength I limit state horizontal shear that
develops at the interface of the slab and girder in a composite section. The dead load used to
calculate the horizontal shear should only consider the DC and DW dead loads that act on the
composite section. See 17.2.4 for further information regarding the treatment of dead loads
and load combinations.
Vni ≥ Vui / φ
Where:
The shear stress at the interface between the slab and the girder is given by:
Vu
v ui =
b vidv
Where:
Vui = 12v ui b vi
[
Vni = cA cv + µ A vf fy + Pc ]
Where:
Pc shall include the weight of the deck, haunch, parapets, and future wearing surface. A
conservative assumption that may be considered is to set Pc = 0.0 .
The nominal interface shear resistance, Vni, shall not exceed the lesser of:
Where:
The stirrups that extend into the deck slab presented on the Standards are considered adequate
to satisfy the minimum reinforcement requirements of LRFD [5.8.4.4]
Web shear reinforcement shall be designed by LRFD [5.8.3.4.3] (Simplified Procedure) using the
Strength I limit state for WisDOT standard girders.
WisDOT prefers girders with spacing symmetrical about the midspan in order to simplify design
and fabrication. The designer is encouraged to simplify the stirrup arrangement as much as
possible. For vertical stirrups, the required area of web shear reinforcement is given by the
following equation:
( Vn − Vc )s b vs
Av ≥ (or 0.0316λ f ' c minimum , LRFD [5.8.2.5])
f y d v cot θ fy
Where:
fpc
• When Vci > Vcw, cot θ = 1.0 + 3 ≤ 1 .8
f'
c
Where:
Per LRFD [5.8.3.4.3], determine Vc as the minimum of either Vci or Vcw given by:
Vi M cre
Vci = 0.02λ f ' c b v d v + Vd + ≥ 0.06λ f ' c b v d
M max
Where:
Vi = Vu − Vd
12Mdnc
Mcre = S c fr + fcpe −
Snc
Mmax = Mu − M dnc
Where:
For a composite section, Vci corresponds to shear at locations of accompanying flexural stress.
Vcw corresponds to shear at simple supports and points of contraflexure. The critical
computation for Vcw is at the centroid for composite girders.
Set the vertical component of the draped strands, Vp, equal to 0.0 when calculating Vn, as per
LRFD [5.8.3.3]. This vertical component helps to reduce the shear on the concrete section.
The actual value of Vp should be used when calculating Vcw. However, the designer may make
the conservative assumption to neglect Vp for all shear resistance calculations.
Based on past performance, for prestressed I-girders the upper limit for web reinforcement
spacing, smax, per LRFD [5.8.2.7] will be reduced to 18 inches.
The web reinforcement spacing shall not exceed the maximum permitted spacing determined
as:
Where:
Vu − φVp
υu = per LRFD [5.8.2.9].
φb v d v
A v f y d v cot θ
Vc + ≤ 0.25 f ' c b v d v
s
Reinforcement in the form of vertical stirrups is required at the extreme ends of the girder. The
stirrups are designed to resist 4% of the total prestressing force at transfer at a unit stress of
20 ksi and are placed within h/4 of the girder end, where h is the total girder depth. For a
distance of 1.5d from the ends of the beams, reinforcement shall be placed to confine the
prestressing steel in the bottom flange. The reinforcement shall be shaped to enclose the
strands, shall be a #3 bar or greater and shall be spaced at less than or equal to 6”. Note that
the reinforcement shown on the Standard Details sheets satisfies these requirements.
Welded wire fabric may be used for the vertical reinforcement. It must be deformed wire with
a minimum size of D18.
Per LRFD [5.8.3.5], at the inside edge of the bearing area to the section of critical shear, the
longitudinal reinforcement on the flexural tension side of the member shall satisfy:
V
A s f y + A ps fps ≥ u − 0.5 Vs cot θ
φ
In the above equation, cot θ is as defined in the Vc discussion above, and Vs is the shear
reinforcement resistance at the section considered. Any lack of full reinforcement development
shall be accounted for. Note that the reinforcement shown on the Standard Detail sheets
satisfies these requirements.
The design of non-prestressed reinforcement for negative moment at the support is based on
the Strength I limit state requirements of LRFD [5.7.3]:
LRFD [5.5.4.2] allows a φ factor equal to 0.9 for tension-controlled reinforced concrete
sections such as the bridge deck.
The continuity reinforcement consists of mild steel reinforcement in the deck in the negative
moment region over the pier. Consider both the non-composite and the superimposed dead
loads and live loads for the Strength I design of the continuity reinforcement in the deck.
Moment resistance is developed in the same manner as shown in 19.3.3.13.1 for positive
moments, except that the bottom girder flange is in compression and the deck is in tension.
The moment resistance is formed by the couple resulting from the compression force in the
bottom flange and the tension force from the longitudinal deck steel. Consider As to consist of
the longitudinal deck steel present in the deck slab effective flange width as determined in
19.3.3.8. The distance, dp, is taken from the bottom of the girder flange to the center of the
longitudinal deck steel.
New bridge designs shall consider only the top mat of longitudinal deck steel when computing the
continuity reinforcement capacity.
The continuity reinforcement shall be based on the greater of either the interior girder design or
exterior girder and detailed as typical reinforcement for the entire width of the bridge deck.
However, do not design the continuity steel based on the exterior girder design beneath a raised
sidewalk. The continuity steel beneath a raised sidewalk should not be used for rating.
Based on the location of the neutral axis, the bottom flange compressive force may behave as
either a rectangle or a T-section. On WisDOT standard prestressed I-girders, if the depth of
the compression block, a, falls within the varying width of the bottom flange, the compression
block acts as an idealized T-section. In this case, the width, b, shall be taken as the bottom
flange width, and the width, bw, shall be taken as the bottom flange width at the depth “a”.
During T-section behavior, the depth, hf, shall be taken as the depth of the bottom flange of full
width, b. See Figure 19.3-4 for details. Ensure that the deck steel is adequate to satisfy
Mr ≥ Mu .
bw = Equivalent width of
bw
web of prestressed
beam for T-sections
hf
Figure 19.3-4
T-Section Compression Flange Behavior
The continuity reinforcement should also be checked to ensure that it meets the crack control
provisions of LRFD [5.7.3.4]. This check shall be performed assuming severe exposure
conditions. Only the superimposed loads shall be considered for the Service and Fatigue
requirements.
The concrete between the abutting girder ends is usually of a much lesser strength than that
of the girders. However, tests1 have shown that, due to lateral confinement of the diaphragm
concrete, the girder itself fails in ultimate negative compression rather than failure in the
material between its ends. Therefore the ultimate compressive stress, f'c, of the girder concrete
is used in place of that of the diaphragm concrete.
This assumption has only a slight effect on the computed amount of reinforcement, but it has
a significant effect on keeping the compression force within the bottom flange.
The continuity reinforcement shall conform to the Fatigue provisions of LRFD [5.5.3].
The transverse spacing of the continuity reinforcement is usually taken as the whole or
fractional spacing of the D bars as given in 17.5.3.2. Grade 60 bar steel is used for continuity
reinforcement. Required development lengths for deformed bars are given in Chapter 9 –
Materials.
The continuity reinforcement is not required to be anchored in regions of the slab that are in
compression at the strength limit state as stated in LRFD [5.14.1.4.8]. The following locations
shall be used as the cut off points for the continuity reinforcement:
1. When ½ the bars satisfy the Strength I moment envelope (considering both the non-composite
and composite loads) as well as the Service and Fatigue moment envelopes (considering only
the composite moments), terminate ½ of the bars. Extend these bars past this cutoff point a
distance not less than the girder depth or 1/16 the clear span for embedment length requirements.
2. Terminate the remaining one-half of the bars an embedment length beyond the point of
inflection. The inflection point shall be located by placing a 1 klf load on the composite structure.
This cut-off point shall be at least 1/20 of the span length or 4’ from point 1, whichever is greater.
Certain secondary features result when spans are made continuous. That is, positive moments
develop over piers due to creep5, shrinkage and the effects of live load and dynamic load
allowance in remote spans. The latter only exists for bridges with three or more spans.
These positive moments are somewhat counteracted by negative moments resulting from
differential shrinkage4 between the cast-in-place deck and precast girders along with negative
moments due to superimposed dead loads. However, recent field observations cited in LRFD
[C5.14.1.4.2] suggest that these moments are less than predicted by analysis. Therefore,
negative moments caused by differential shrinkage should be ignored in design.
WisDOT requires the use of a negative moment connection only. The details for a positive
moment connection per LRFD [5.14.1.4] are not compatible with the Standard Details and should
not be provided.
The prestress camber and dead load deflection are used to establish the vertical position of
the deck forms with respect to the girder. The theory presented in the following sections apply
to a narrow set of circumstances. The designer is responsible for ensuring that the theoretical
camber accounts for the loads applied to the girder. For example, if the diaphragms of a
prestressed I-girder are configured so there is one at each of the third points instead of one at
midspan, the term in the equation for ∆ nc (DL ) related to the diaphragms in 19.3.3.18.2 would
need to be modified to account for two point loads applied at the third points instead of one
point load applied at midspan.
Deflection effects due to individual loads may be calculated separately and superimposed, as
shown in this section. The PCI Design Handbook provides design aids to assist the designer
in the evaluation of camber and deflection, including cambers for prestress forces and loads,
and beam design equations and diagrams.
Figure 19.3-5 illustrates a typical prestressed I-girder with a draped strand profile.
Figure 19.3-5
Typical Draped Strand Profile
The prestressing strands produce moments in the girder as a result of their eccentricity and
draped pattern. These moments induce a camber in the girder. The values of the camber are
calculated as follows:
M1 = (Pi ( y B − yy))
1 s
12
Where:
M1 = Moment due to initial prestress force in the straight strands minus the
elastic shortening loss (k-ft)
Pi s = Initial prestress force in the straight strands minus the elastic
shortening loss (kips)
yB = Distance from center of gravity of beam to bottom of beam (in)
M1L2
∆s = (with all units in inches and kips)
8E iIb
For moments expressed in kip-feet and span lengths expressed in feet, this equation becomes
the following:
216M1L2
∆s = (with units as shown below)
E iIb
Where:
The draped strands induce the following moments at the ends and within the span:
Where:
These moments produce a net upward deflection at midspan, which is given by:
216L2 23
∆D = M 2 − M3
E iIb 27
Where:
216L2 23
∆ PS = ∆ s + ∆ D = M1 + M 2 − M3
E iIb 27
Where:
5 WbL4
∆ o (DL ) = (with all units in inches and kips)
384E iIb
Using unit weights in kip per foot, span lengths in feet, E in ksi and Ib in inches4, this equation
becomes the following:
22.5 WbL4
∆ o (DL ) = (with units as shown below)
E iIb
Where:
∆ i = ∆ PS − ∆ o (DL )
Where:
Camber, however, continues to grow after the initial strand release. For determining
substructure beam seats, average concrete haunch values (used for both DL and quantity
calculations) and the required projection of the vertical reinforcement from the tops of the
prestressed girders, a camber multiplier of 1.4 shall be used. This value is multiplied by the
theoretical camber at release value.
The downward deflection of a prestressed I-girder due to the dead load of the deck and a
midspan diaphragm is:
5 W deck L4 Pdia L3
∆ nc (DL ) = + (with all units in inches and kips)
384EIb 48EIb
Using span lengths in units of feet, unit weights in kips per foot, E in ksi, and Ib in inches4, this
equation becomes the following:
22.5 Wb L4 36Pdia L3
∆ o (DL ) = + (with units as shown below)
EIb EIb
Where:
A similar calculation is done for parapet and sidewalk loads on the composite section.
Provisions for deflections due to future wearing surface shall not be included.
For girder structures with raised sidewalks, loads shall be distributed as specified in Chapter
17, and separate deflection calculations shall be performed for the interior and exterior girders.
Residual camber is the camber that remains after the prestress camber has been reduced by
the composite and non-composite dead load deflection. Residual camber is computed as
follows:
RC = ∆ i − ∆ nc ( DL ) − ∆ c ( DL )
Deck forming requires computing the relationship between the top of girder and bottom of deck
necessary to achieve the desired vertical roadway alignment. Current practice for design is to
use a minimum haunch of 2" at the edge of the girder flange. This haunch value is also used
for calculating composite section properties. This will facilitate current deck forming practices
which use 1/2" removable hangers and 3/4" plywood, and it will allow for variations in prestress
camber. Also, future deck removal will be less likely to damage the top girder flanges. An
average haunch height of 3 inches minimum can be used for determining haunch weight for
preliminary design. It should be noted that the actual haunch values should be compared with
the estimated values during final design. If there are significant differences in these values,
the design should be revised. The actual average haunch height should be used to calculate
the concrete quantity reported on the plans as well as the value reported on the prestressed
girder details sheet. The actual haunch values at the girder ends shall be used for determining
beam seat elevations.
For designs involving vertical curves, Figure 19.3-6 shows two different cases.
Figure 19.3-6
Relationship between Top of Girder and Bottom of Deck
In Case (a), VC is less than the computed residual camber, RC, and the minimum haunch
occurs at midspan. In Case (b), VC is greater than RC and the minimum haunch occurs at the
girder ends.
Deck forms are set to accommodate the difference between the bottom of the deck and the
top of the girder under all dead loads placed at the time of construction, including the wet deck
concrete and superimposed parapet and sidewalk loads. The deflection of superimposed
future wearing surface and live loads are not included.
For equal-span continuous structures having all spans on the same vertical alignment, the
deck forming is the same for each span. This is due to the constant change of slope of the
vertical curve or tangent and the same RC per span.
+ HEND = RC − VC + ( +HCL )
Where:
For unequal spans or when some spans are on a vertical curve and others are on a tangent,
a different approach is required. Generally the longer span or the one off the curve dictates the
haunch required at the common support. Therefore, it is necessary to pivot the girder about its
midspan in order to achieve an equal condition at the common support. This is done
mathematically by adding together the equation for each end (abutment and pier), as follows:
Where:
With the condition at one end known due to the adjacent span, the condition at the other end
is computed.
The transverse construction joints should be located in the deck midway between the cut-off
points of the continuity reinforcement or at the 0.75 point of the span, whichever is closest to
the pier. The construction joint should be located at least 1' from the cut-off points.
This criteria keeps stresses in the slab reinforcement due to slab dead load at a minimum and
makes deflections from slab dead load closer to the theoretical value.
Low relaxation strands (0.5” and 0.6” in diameter) are currently used in prestressed I-girder
and prestressed box girder designs and are shown on the plans. Strand patterns and initial
prestressing forces are given on the plans, and deflection data is also shown.
Refer to the AASHTO LRFD Bridge Construction Specifications for the required dimensional
tolerances.
WisDOT BOS employs two prestress I-girder section families. One I section family follows the
AASHTO standard section, while the other section family follows a wide flange bulb-tee, see
Figure 19.3-7. These sections employ draped strand patterns with undraped alternates where
feasible. Undraped strand patterns, when practical, should be specified on the designs. For
these sections, the cost of draping far exceeds savings in strands. See the Standard Details
for the prestressed I-girder sections’ draped and undraped strand patterns. Note, for the 28”
prestressed I-girder section the 16 and 18 strand patterns require bond breakers.
Figure 19.3-7
Prestressed I-Girder Family Details
Table 19.3-1 and Table 19.3-2 provide span lengths versus interior girder spacings for HL-93
live loading on single-span and multiple-span structures for prestressed I-girder sections.
Girder spacings are based on using low relaxation strands at 0.75fpu, concrete haunch
thicknesses, slab thicknesses from Chapter 17 – Superstructures - General and a future
wearing surface. For these tables, a line load of 0.300 klf is applied to the girder to account for
superimposed dead loads.
Several girder shapes have been retired from standard use on new structures. These include
the following sizes; 45-inch, 54-inch and 70-inch. These girder shapes are used for girder
replacements, widening and for curved new structures where the wide flange sections are not
practical. See Chapter 40 – Bridge Rehabilitation for additional information on these girder
shapes.
Due to the wide flanges on the 54W, 72W and 82W and the variability of residual camber,
haunch heights frequently exceed 2”. An average haunch of 4” was used for all wide flange
girders in the following tables. Do not push the span limits/girder spacing during
preliminary design. See Table 19.3-2 for guidance regarding use of excessively long
prestressed I-girders.
• Interior prestressed I-girders, 0.5” or 0.6” dia. strands (in accordance with the Standard
Details).
Table 19.3-1
Maximum Span Length vs. Girder Spacing
Table 19.3-2
Maximum Span Length vs. Girder Spacing
* For lateral stability during lifting these girder lengths may require pick-up point locations
greater than distance d (girder depth) from the ends of the girder. The designer shall assume
that the pick-up points will be at the 1/10 points from the end of the girder and provide extra
non-prestressed steel in the top flange, if required, and check the concrete strength near the
lift location based on f’ci. A note should be placed on the girder details sheet to reflect that the
girder was analyzed for a potential lift at the 1/10 point.
The standard strand patterns presented in the Standard Details were developed to eliminate
some of the trial and error involved in the strand pattern selection process. These standard
strand patterns should be used whenever possible, with a straight strand arrangement
preferred over a draped strand arrangement. The designer is responsible for ensuring that the
selected strand pattern meets all LRFD requirements.
Section 19.3.3 discusses the key parts of the design procedure, and how to effectively use the
standard strand patterns along with Table 19.3-1 and Table 19.3-2.
The amount of drape allowed is controlled by the girder size and the 2" clearance from center
of strand to top of girder. See the appropriate Standard Girder Details for guidance on draping.
These sections may be used for skews up to 30° with the transverse post-tensioning ducts
placed along the skew. Skews over 30° are not recommended, but if absolutely required the
transverse post-tensioning ducts should be placed perpendicular to the prestressed sections.
Also for skews over 30° a more refined method of analysis should be used such as an
equivalent plate analysis or a finite element analysis.
Details for transverse post-tensioning are shown in the Standard Details. Each post-tensioning
duct contains three ½” diameter strands which produce a total post-tensioning force per duct
of 86.7 kips.
Prestressed box girders are subject to high chloride ion exposure because of longitudinal
cracking that sometimes occurs between the boxes or from drainage on the fascia girders
when an open steel railing system is used. To reduce permeability the concrete mix is required
to contain fly ash as stated in 503.2.2 of the Standard Specifications.
When these sections are in contact with water for 5-year flood events or less, the sections
must be cast solid for long term durability. When these sections are in contact with water for
the 100-year flood event or less, any voids in the section must be cast with a non-water-
absorbing material.
Table 19.3-3 provides approximate span limitations for prestressed box girder sections. It also
gives the section properties associated with these members. Criteria for developing these
tables are shown below Table 19.3-3.
January 2017 19-45
WisDOT Bridge Manual Chapter 19 – Prestressed Concrete
19.3.9.1 Available Prestressed Box Girder Sections and Maximum Span Lengths
Precasters have forms available to make six prestressed girder box sections ranging in depth
from 12” to 42”. Each section can be made in widths of 36” and 48”, but 48” is more efficient
and is the preferred width. Typical box section information is shown in the Standard Details.
Table 19.3-3 shows available section depths, section properties, and maximum span length.
All sections have voids except the 12” deep section.
Table 19.3-3
Prestressed Box Girder Section Properties and Maximum Span Length
• Simple span
• Single slope parapet 42SS weight distributed evenly to all girder sections
• 30’-0” minimum clear bridge width (eleven 3’-0” sections, eight 4’-0” sections)
These sections are set 1” apart with a ±¼” tolerance. The space between sections is filled with
a grout mix prior to post-tensioning the sections transversely. Post-tensioning is not allowed
until the grout has cured for at least 48 hours and has attained a compressive strength of
3000psi.
When strands are tensioned in open or unheated areas during cold weather they are subject
to loss due to change in temperature. This loss can be compensated for by noting the change
in temperature of the strands between tensioning and initial set of the concrete. For purposes
of uniformity the strand temperature at initial concrete set is taken as 80°F.
Minor changes in temperature have negligible effects on the prestress force, therefore only at
strand temperatures of 50°F and lower are increases in the tensioning force made.
Since plan prestress forces are based on 75% of the ultimate for low relaxation strands it is
necessary to utilize the AASHTO allowable of temporarily overstressing up to 80% to provide
for the losses associated with fabrication.
The following example outlines these losses and shows the elongation computations which
are used in conjunction with jack pressure gages to control the tensioning of the strands.
Known:
Ep = 29,000 ksi (of prestressing tendons, sample tested from each spool)
COMPUTE:
19.5 References
1. Whitney, C. S., "Plastic Theory of Reinforced Concrete Design", ASCE Trans., 107, 1942,
p. 251.
2. Karr, P. H., Kriz, L. B. and Hognestad, E., "Precast-Prestressed Concrete Bridges 1. Pilot
Tests of Continuous Beams", Portland Cement Association Development Department, Bulletin
D34.
4. Freyermuth, Clifford L., "Design of Continuous Highway Bridges with Precast, Prestressed
Concrete Girders (EB014.01E)”, Portland Cement Association, 1969.
5. Lin, T. Y. and Burns, N. H., "Design of Prestressed Concrete Structures", Third Edition, J.
Wiley, 1981.
Table of Contents
E19-1 Single Span Bridge, 72W" Prestressed Girders LRFD...................................................................... 2
E19-1.1 Design Criteria ...................................................................................................... 2
E19-1.2 Modulus of Elasticity of Beam and Deck Material................................................. 3
E19-1.3 Section Properties ................................................................................................ 3
E19-1.4 Girder Layout ........................................................................................................ 4
E19-1.5 Loads .................................................................................................................... 4
E19-1.5.1 Dead Loads .......................................................................................... 5
E19-1.5.2 Live Loads ............................................................................................ 5
E19-1.6 Load Distribution to Girders .................................................................................. 6
E19-1.6.1 Distribution Factors for Interior Beams: ................................................ 7
E19-1.6.2 Distribution Factors for Exterior Beams: ............................................... 7
E19-1.6.3 Distribution Factors for Fatigue:............................................................ 9
E19-1.7 Limit States and Combinations ............................................................................. 9
E19-1.7.1 Load Factors......................................................................................... 9
E19-1.7.2 Dead Load Moments ............................................................................10
E19-1.7.3 Live Load Moments ..............................................................................10
E19-1.7.4 Factored Moments................................................................................11
E19-1.8 Composite Girder Section Properties ...................................................................12
E19-1.9 Preliminary Design Information:............................................................................13
E19-1.10 Preliminary Design Steps ...................................................................................16
E19-1.10.1 Determine Amount of Prestress..........................................................16
E19-1.10.2 Prestress Loss Calculations ...............................................................18
E19-1.10.2.1 Elastic Shortening Loss ......................................................18
E19-1.10.2.2 Approximate Estimate of Time Dependant Losses.............19
E19-1.10.3 Design of Strand Drape ......................................................................20
E19-1.10.4 Stress Checks at Critical Sections ......................................................26
E19-1.11 Calculate Jacking Stress ....................................................................................31
E19-1.12 Flexural Strength Capacity at Midspan...............................................................32
E19-1.13 Shear Analysis....................................................................................................36
E19-1.14 Longitudinal Tension Flange Capacity:...............................................................44
E19-1.15 Composite Action Design for Interface Shear Transfer ......................................45
E19-1.16 Deflection Calculations .......................................................................................47
E19-1.17 Camber Calculations ..........................................................................................47
WisDOT Bridge Manual Chapter 19 – Prestressed Concrete
40'-0" Clear
f' ci 6.8 girder initial concrete strength, ksi New limit for release strength.
ts 8 slab thickness, in
wc 0.150 kcf
ED Edeck4
EB
n
ED n 1.540
Note that this value of EB is used for strength, composite section property, and long term
deflection (deck and live load) calculations.
The value of the modulus of elasticity at the time of release is calculated in accordance with
LRFD [C5.4.2.4]. This value of Ect is used for loss and instantaneous deflection (due to
prestress and dead load of the girder) calculations.
1.5
Ebeam6.8 33000 wc f'ci Ebeam6.8 4999
Ect Ebeam6.8
tt 5.5 in
tw 6.5 in
tw
tb 13 in
ht 72 in tb
bw 30 width of bottom flange, in
Ag 915 in2
tse
eg yt 2 eg 42.88 in
2
Kg n Ig Ag eg
2
Kg 3600866 in4
LRFD [Eq 4.6.2.2.1-1]
Chapter 19 suggests that at a 146 foot span, the girder spacing should be 8'-6" with 72W
girders.
S 8.5 ft
Assume a minimum overhang of 2.5 feet (2 ft flange + 6" overhang), soh 2.5
wb 2 soh
nspa nspa 4.412
S
Use the next lowest integer: nspa ceil nspa nspa 5
wb 2 soh
S S 7.50 ft
nspa
E19-1.5 Loads
wg 0.953 weight of 72W girders, klf
w wws
wws wws 0.133 klf
ng
* LRFD [4.6.2.2.1] states that permanent loads on the deck may be distributed uniformly
among the beams. This method is used for the parapet and future wearing surface loads.
For Fatigue:
LRFD [5.5.3] states that fatigue of the reinforcement need not be checked for fully
prestressed components designed to have extreme fiber tensile stress due to Service III
Limit State within the tensile stress limit specified in LRFD [Table 5.9.4.2.2-1].
For fully prestressed components, the compressive stress due to the Fatigue I load
combination and one half the sum of effective prestress and permanent loads shall not
exceed 0.40 f'c after losses.
Distribution factors are in accordance with LRFD [Table 4.6.2.2.2b-1]. For an interior beam,
the distribution factors are shown below:
Criteria for using distribution factors - Range of Applicability per LRFD [Table 4.6.2.2.2b-1].
"NG" otherwise
NoBeams "OK" if ng 4
"NG" otherwise
"NG" otherwise
S DeckSpan
7.5 "OK"
tse DeckThickness 7.5 "OK"
x L BridgeSpan x 146.0 "OK"
ng NoBeams 6.0 "OK"
K LongitStiffness
g 3600866.5 "OK"
gi max gi1 gi2 gi 0.636
Note: The distribution factors above already have a multiple presence factor included that is
used for service and strength limit states. For fatigue limit states, the 1.2 multiple presence
factor should be divided out.
Per LRFD [Table 4.6.2.2.2d-1] the distribution factor shall be calculated by the following
equations:
wb w
wparapet Width of parapet wparapet 1.250 ft
2 overlapping the deck
"NG" otherwise
Note: While AASHTO allows the de value to be up to 5.5, the deck overhang (from the
center of the exterior girder to the edge of the deck) is limited by WisDOT policy as stated
in Chapter 17 of the Bridge Manual.
de
e 0.77 e 0.907
9.1
sw1 sw2
2'-0" 6'-0"
de S
S sw1 sw2
Rx Rx 0.500 % of a lane load
S 2
Add the single lane multi-presence factor, m 1.2
The exterior girder distribution factor is the maximum value of the One Lane Loaded case and
the Two or More Lanes Loaded case:
gx max gx1 gx2 gx 0.600
Note: The interior girder has a larger live load distribution factor and a larger dead load than
the exterior girder. Therefore, for this example, the interior girder is likely to control.
gi1
gif gif 0.362
1.2
The limit states, load factors and load combinations shall be applied as required and detailed in
chapter 17 of this manual and as indicated below.
The DCnc values are the component non-composite dead loads and include the weight of the
girder, haunch, diaphragms and the deck.
The DCc values are the component composite dead loads and include the weight of the
parapets.
The DWc values are the composite dead loads from the future wearing surface.
Note that the girder dead load moments at release are calculated based on the girder length.
The moments for other loading conditions are calculated based on the span length (center to
center of bearing).
The Wisconsin Standard Permit Vehicle should also be checked. See Chapter 45 - Bridge
Rating for further information.
The unfactored live load moments per lane are calculated by applying the appropriate
distribution factor to the controlling moment. For the interior girder:
gi 0.636
gif 0.362
Mstr η γstDC MDLnc MDLc γstDW MDWc γstLL MLL
1.0 1.25 MDLnc MDLc 1.50 MDWc 1.75 MLL Mstr 12449 kip-ft
Ms1 η γs1DC MDLnc MDLc γs1DW MDWc γs1LL MLL
1.0 1.0 MDLnc MDLc 1.0 MDWc 1.0 MLL Ms1 8659 kip-ft
Ms3 η γs3DC MDLnc MDLc γs3DW MDWc γs3LL MLL
1.0 1.0 MDLnc MDLc 1.0 MDWc 0.8 MLL Ms3 8045 kip-ft
Fatigue 1
Mfat η γfLL MLLfat Mfat 1307 kip-ft
we S 12 we 90.00 in
The effective width, we , must be adjusted by the modular ratio, n, to convert to the same
concrete material (modulus) as the girder.
we
weadj weadj 58.46 in
n
hc 81.50 in
n 1.540
Note: The area of the concrete haunch is not included in the calculation of the composite
section properties.
ΣA 1353 in2
ΣAY
ycgb ycgb 48.8 in
ΣA
2
Icg ΣIplusAYsq Acg ycgb Icg 1203475 in4
Icg
Scgt Scgt 51786 in3
ycgt
Icg
Scgb Scgb 24681 in3
ycgb
Deck:
Icg
Scgdt n in3
ycgt hau tse Scgdt 56594
Icg
Scgdb n Scgdb 73411 in3
ycgt hau
Note: The initial girder loads will be used to check stresses at transfer (before losses) and the
service loads will be used to check stresses while in service (after losses). These calculations
and the estimated prestress losses will then be used to select the number of strands for final
design calculations.
Miend 0 kip-ft
2
Lg
Mg wg Mg 2574 kip-ft
8
Note: The estimated time dependant losses (for low relaxation strands) will be
re-calculated using the approximate method in accordance with LRFD [5.9.5.3] once th
number of strands has been determined.
Assume an initial strand stress; ftr 0.75 fpu ftr 202.500 ksi
ΔfpES_est
ESloss 100 ESloss 8.889 %
ftr
The total loss is the time dependant losses plus the ES losses:
loss
loss% 100 loss% 24.7 % (estimated)
ftr
T = 1 loss% To
loss%
ratio 1 ratio 0.753
100
T = ratio To
1) Design the amount of prestress to prevent tension at the bottom of the beam under the full
| load at center span after losses.
2) Calculate the prestress losses and check the girder stresses at mid span at the time of
transfer.
3) Design the eccentricity of the strands at the girder end to avoid tension or compression
over-stress at the time of transfer.
5) Check resulting stresses at the critical sections of the girder at the time of transfer and after
| losses.
Near center span, after losses, T = the remaining effective prestress, aim for no tension at the
| bottom. Use Service I for compression and Service III for tension.
For this example, the interior girder has the controlling moments.
Calculate the stress at the bottom of the beam due to combination of non-composite and
composite loading (Service 3 condition):
Mnc 12 M3c 12
fb fb 4.651 ksi
Sb Scgb
Stress at bottom due to prestressing (after losses):
T yb
fbp =
A
1 e
2
where T = 1 loss% T o
r
and fbp fb desired final prestress.
We want this to balance out the tensile stress calculated above from the loading, i.e. an initial
compression. Since we are making some assumptions on the actual losses, we are ignoring
the allowable tensile stress in the concrete for these calculations.
fbp To yb
= 1 e
1 loss% A 2
r
fbp
fbpi_1
loss% fbpi_1 6.175 ksi
1
100
desired bottom initial prestress (before losses)
If we use the actual allowable tensile stress in the concrete, the desired bottom initial prestress
is calculated as follows:
ftall 0.19 λ f' c < 0.6 ksi; λ = 1.0 (norm. wgt. conc.) LRFD [5.4.2.8] ftall 0.537 ksi
Determine the stress effects for different strand patterns on the 72W girder:
As 0.217 in2
fs
P As kips
1000 P 43.943
P N yb
fbpi 1 e (bottom initial prestress - before losses)
Ag rsq
The values of fbpi for various strand patterns is shown in the following table.
72W Stress Effects
Pi (per strand) = 43.94 kips
bottom stress
No. Strands e (in) (ksi)
36 -31.09 4.3411
38 -30.98 4.5726
40 -30.87 4.8030
42 -30.77 5.0333
44 -30.69 5.2648
46 -30.52 5.4858
48 -30.37 5.7075
50 -30.23 5.9290
52 -30.10 6.1504
Solution:
Try ns 46 strands, 0.6 inch diameter.
Initial prestress at bottom fbpi 5.4858 ksi,
Eccentricity, es 30.52 inches; actual tension should be less than allowed.
3) Creep (CR), slow shortening of concrete due to permanent compression stresses in the
beam, time function.
4) Relaxation (RE), the tendon slowly accommodates itself to the stretch and the internal
stress drops with time
The ES loss estimated above was: ΔfpES_est 18.0 ksi, or ESloss 8.889 %. The
resulting force in the strands after ES loss:
ESloss
T o 1 Toi T o 1842 kips
100
If we assume all strands are straight we can calculate the initial elastic shortening loss;
To es es
fcgp
Ag
T o es
Ig
Mg 12
Ig
fcgp 3.190 ksi
Ep Es Ep 28500 ksi
Ep
ΔfpES f ΔfpES 18.185 ksi
Ect cgp
This value of fpES is in agreement with the estimated value above; ΔfpES_est 18.00 ksi. If
these values did not agree, T o would have to be recalculated using ftr minus the new value of
fpES, and a new value of fcgp would be determined. This iteration would continue until the
assumed and calculated values of fpES are in agreement.
Check the design to avoid premature failure at the center of the span at the time of transfer.
Check the stress at the center span (at the plant) at both the top and bottom of the girder.
To T o es Mg 12
fttr fttr 0.582 ksi
Ag St St
To T o es Mg 12
fbtr fbtr 3.353 ksi
Ag Sb Sb
Is the stress at the bottom of the girder less than the allowable? check "OK"
fpi As
ΔfpLT = 10.0 γh γst 12.0 γh γst Δf
Ag pR
From LRFD [Figure 5.4.2.3.3-1], the average annual ambient relative humidity, H 72 %.
5
γst γst 0.641
1 f' ci
ftr As ns
ΔfpCR 10.0 γh γst ΔfpCR 13.878 ksi
Ag
Δfp
100 20.741 % total
ftr prestress loss
This value is slightly less than but in general agreement with the initial estimated
loss% 24.691 .
The remaining stress in the strands and total force in the beam after all losses is:
top:
To T o es
ftetr ftetr 1.165 ksi
Ag St
high tension stress
In accordance with LRFD Table [5.9.4.1.2-1], the temporary allowable tension stress is
calculated as follows (assume there is no bonded reinforcement):
ftiall min 0.0948 λ f' ci 0.2 λ = 1.0 (normal wgt. conc.)
LRFD [5.4.2.8]
ftiall 0.200 ksi
bottom:
To T o es
fbetr fbetr 4.994 ksi
Ag Sb
fciall 4.420 ksi
Drape some of the strands upward to decrease the top tension and decrease the compression
at the bottom.
Find the required position of the steel centroid to avoid tension at the top. Conservatively set
the top stress equal to zero and solve for "e":
To T o es
ftetr =
Ag St
St To
esendt 0 esendt 19.32 inches
To Ag or higher
Find the required position of the steel centroid to avoid high compression at the bottom of the
beam. Set the bottom compression equal to the allowable stress and find where the centroid
of ns 46 strands needs to be:
To T o es
fbetr =
Ag Sb
Sb To
esendb fciall esendb 24.65 inches
To Ag or higher
esend max esendt esendb esend 19.32 inches
cgc
es = 30.52"
yb = 34.87"
LRFD [Table 5.12.3-1] requires 2 inches of cover. However, WisDOT uses 2 inches to the
center of the strand, and 2 inch spacing between centers.
The center nsd 8 strands will be draped at the end of the girder.
Find the center of gravity of the remaining nss 38 straight strands from the bottom of the
girder:
12 2 12 4 12 6 2 8
Ys Ys 4.21 inches from the
nss bottom of the girder
OR:
y8t cgc
ys straight
strands
y8t is the eccentricity of the draped strands at the end of the beam. We want the eccentricity of
all of the strands at the end of the girder to equal, esend 19.322 inches for stress
control.
However, yt 37.13 inches to the top of the beam. If the draped strands are raised
y8t 34.53 inches or more above the cgc, the stress will be OK.
Drape the center strands the maximum amount: Maximum drape for nsd 8 strands:
es 30.52 in
y8t es 62.65 in
draped
strands
y 8t cgc
y 8t - e s
ys straight es
strands
L/3 or 0.4*L
Lg Lg
Try a drape length of: 49.00 feet HD
3 3
The eccentricity of the draped strands at the hold down point:
e8hd yb 5 e8hd 29.870 in
y8t e8hd
Strand slope, slope 100
( HD 12 ) slope 10.54 %
12% is a suggested maximum slope, actual acceptable slope is dependant on the form capacity
or on the fabricator.
Center of Gravity
of Draped Strands
Hold Down
End of Point
A Girder
Bottom C
of Girder B
¼ point (0.25L)
A yb y8t A 67.00 in
C 5.00 in
A 3C
Bmin Bmin 20.50 in
4
Check hold down location for Bmax to make sure it is located between Lg /3 and 0.4*Lg :
A Bmax
slopeBmax slopeBmax 0.099 ft/ft
0.25 Lg 12
AC 1
xBmax xBmax 52.38 ft
slopeBmax 12
Is the resulting hold down location less than 0.4*Lg ? check "OK"
Check the girder stresses at the end of the transfer length of the strands at release:
ltr
x x 3.00 feet
12
The eccentricity of the draped strands and the entire strand group at the transfer length is:
slope
y8tt y8t x 12 y8tt 28.334 in
100
nss ys 8 y8tt
est est 20.400 in
ns
The moment at the end of the transfer length due to the girder dead load:
wg
L x x
2
Mgt Mgt 206 kip-ft
2 g
To T o est Mgt 12
fbt fbt 3.873 ksi
Ag Sb Sb
fciall 4.420 ksi
Lg
Need moments at hold down points: 49.00 feet, from the end of the girder.
3
girder: Mghd 2288 kip-ft
Note: The release girder moments shown above at the hold down location are calculated based
on the total girder length.
Check the girder at the end of the beam (at the transfer length):
es 30.52 inches
Final condition:
tension: ftall = 0.19 λ f' c λ = 1.0 (normal wgt. conc.) LRFD [5.4.2.8]
ftall 0.19 f' c | ftall | < 0.6 ksi ftall 0.537 ksi
T T es MDLnc 12
MDLc MDWc 12 ksi
ft1 ft1 2.465
Ag St St Scgt
T T es MDLnc 12
MDLc MDWc MLL 12 ksi
ft2 ft2 3.177
Ag St St Scgt
1 M
T es MDLnc 12 DLc MDWc MLLfat 12
1 T 2
ftfat
2 Ag St St Scgt
ftfat 1.434 ksi
T T es Mnc 12
Ms3 Mnc 12 ksi
fb fb 0.302
Ag Sb Sb Scgb
| Final condition, after losses, full load: ftall 0.537 ksi fcall2 4.800 ksi
T T es Mnchd 12 M1chd 12
ft3 ft3 2.710 ksi
Ag St St Scgt
Is ft3 less than fcall2 ? check "OK"
1 T T es Mnchd 12 Mfatchd 12
ftfat ftfat 1.317 ksi
2 Ag St St Scgt
Is ftfat less than fcall_fat? check "OK"
T T es Mnchd 12 M3chd 12
fb3 fb3 0.212 ksi
Ag Sb Sb Scgb
Is fb3 greater than ftall? check "OK"
M1chd 12 ksi
fdt3 fdt3 0.715
Scgdt
Is fdt less than fdall? check "OK"
M1chd 12
fdb3 fdb3 0.551 ksi
Scgdb
Is fdb less than fdall? check "OK"
Final Allowable:
compression1 fcall1 3.6 ksi
fpy
k = 2 1.04
fpu
From LRFD Table [C5.7.3.1.1-1], for low relaxation strands, k 0.28 .
"c" is defined as the distance between the neutral axis and the compression face (inches).
Assumed dimensions:
tse
hau
yt
tw
es
Assume that the compression block is in the deck. Calculate the capacity as if it is a
rectangular section (with the compression block in the flange). The neutral axis
location,calculated in accordance with LRFD 5.7.3.1.1 for a rectangular section, is:
Aps fpu
c=
fpu
α1 f'cd β1 b k Aps
dp
where:
Aps ns As Aps 9.98 in2
b we b 90.00 in
β1 max 0.85 f' cd 4 0.05 0.65 β1 0.850
a β1 c a 8.49 in
The calculated value of "a" is greater than the deck thickness. Therefore, the rectangular
assumption is incorrect and the compression block extends into the haunch. Calculate the
neutral axis location and capacity for a flanged section:
Aps fpu α1 f'cd b wtf hf in
c c 10.937
fpu
α1 f' cd β1 wtf k Aps
dp
a β1 c a 9.30 in
This is within the depth of the haunch (9.5 inches). Therfore our assumption is OK.
c
fps fpu 1 k
dp fps 259.283 ksi
T u fps Aps T u 2588 kips
Calculate the nominal moment capacity of the composite section in accordance with LRFD
[5.7.3.2]; [5.7.3.2.2]
a a hf 1
2
Mn Aps fps dp α1 f'cd b wtf hf
2 2 12
Mn 15717 kip-ft
For prestressed concrete, ϕf 1.00, LRFD [5.5.4.2.1]. Therefore the usable capacity is:
Mr ϕf Mn Mr 15717
kip-ft
The required capacity:
Interior Girder Moment Mstr 12449 kip-ft
Check the section for minimum reinforcement in accordance with LRFD [5.7.3.3.2] for the
interior girder:
1.33 Mstr 16558 kip-ft
fr 0.24 f'c λ = 1.0 (normal wgt. conc.) LRFD [5.4.2.8] fr 0.679 ksi
T T es
fcpe fcpe 4.348 ksi
Ag Sb
1 Sc
Mcr γ3 Sc γ1 fr γ2 fcpe
12
Mdnc 1 Mcr 10551 kip-ft
Snc
Is Mr greater than the lesser value of Mcr and 1.33*Mstr? check "OK"
The moment capacity looks good, with some over strength for the interior girder. However, we
must check the capacity of the exterior girder since the available flange width is less.
The effective flange width for exterior girder is calculated in accordance with LRFD [4.6.2.6] as
one half the effective width of the adjacent interior girder plus the overhang width :
wex_oh soh 12
wex_oh 30.0 in
we
wex wex_oh in
2 wex 75.00
Aps fpu α1 f' cd bx wtf hf
cx in
fpu cx 13.51
α1 f'cd β1 wtf k Aps
dp
ax β1 cx ax 11.49 in
cx
fps_x fpu 1 k fps_x 256.759 ksi
dp
The nominal moment capacity of the composite section (exterior girder) ignoring the increased
strength of the concrete in the girder flange:
ax ax hf 1
Mn_x Aps fps dp
2
α1 f'cd bx wtf hf
2 2 12
Since Mr_x is greater than 1.33*Mstrx, the check for Mcr does not need to be completed.
Interior Beams:
One lane loaded:
S
gvi1 0.36 gvi1 0.660
25
Two or more lanes loaded:
2
gvi2 0.2
S S
gvi2 0.779
12 35
gvi max gvi1 gvi2 gvi 0.779
Note:The distribution factors above include the multiple lane factor. The skew correction
factor, as now required by a WisDOT policy item for all girders, is omitted. This example
is not yet revised.
Exterior Beams:
Two or more lanes loaded:
The distance from the centerline of the exterior beam to the inside edge of the parapet,
de 1.25 feet.
de
ev 0.6 ev 0.725
10
With a single lane loaded, we use the lever rule (same as before). Note that the multiple
presence factor has already been applied to gx2 ..
gvx max gvx1 gvx2 gvx 0.600
0.3
12L t 3
s π
skewcorrection 1.0 0.2 tan skew
Kg 180
L 146.00
ts 8.00
Kg 3600866
skewcorrection 1.048
skew 20.000
The interior girder will control. It has a larger distribution factor and a larger dead load.
Conduct a bridge analysis as before with similar load cases for the maximum girder shear
forces. We are interested in the Strength 1 condition now for shear design.
300
(kips)
Vu 200
100
0
0 0.1 0.2 0.3 0.4 0.5
TenthPoints Vu0.0 379.7 kips
bv tw bv 6.50 in
The critical section for shear is taken at a distance of dv from the face of the support, LRFD
[5.8.3.2].
dv = effective shear depth taken as the distance between the resultants of the tensile and
compressive forces due to flexure. It need not be taken less than the greater of 0.9*de or
0.72h (inches). LRFD [5.8.2.9]
a
dv es yt hau tse dv 72.50 in
2
However, since there are draped strands for a distance of HD 49.00 feet from the end of
the girder, a revised value of es should be calculated based on the estimated location of the
critical section. Since the draped strands will raise the center of gravity of the strand group
near the girder end, try a smaller value of "dv" and recalculate "es" and "a".
Try dv 65 inches.
For the standard bearing pad of width, wbrg 8 inches, the distance from the end of the
girder to the critical section:
wbrg 1
Lcrit dv 0.5
Lcrit 6.25 ft
2 12
Calculate the eccentricity of the strand group at the critical section.
slope
y8t_crit y8t Lcrit 12 y8t_crit 24.22 in
100
Aps_crit nsd nss As Aps_crit 9.98 in2
Also, the value of fpu , should be revised if the critical section is located less than the
development length from the end of the beam. The development length for a prestressing
strand is calculated in accordance with LRFD [5.11.4.2]:
db 0.600 in
2
ld K fps fpe db ld 146.2 in
3
The transfer length may be taken as: ltr 60 db ltr 36.00 in
Since Lcrit 6.250 feet is between the transfer length and the development length, the
design stress in the prestressing strand is calculated as follows:
Lcrit 12 ltr
fpu_crit fpe
ld ltr
fps fpe fpu_crit 195 ksi
Aps_crit fpu_crit
c in
fpu_crit c 7.276
α1 f' cd β1 b k Aps_crit
dp_crit
acrit
dv_crit es_crit yt hau tse dv_crit 64.65 in
2
This value matches the assumed
value of dv above. OK!
The nominal shear resistance of the section is calculated as follows, LRFD [5.8.3.3]:
Vn = min Vc Vs Vp 0.25 f'c bv dv Vp
Note, the value of Vp does not equal zero in the calculation of Vcw.
Vd = shear force at section due to unfactored dead load and includes both DC and DW (kips)
Vi = factored shear force at section due to externally applied loads (Live Loads) occurring
simultaneously with Mmax (kips). (Not necessarily equal to Vu .)
Mcre = moment causing flexural cracking at section due to externally applied loads (kip-in)
Mmax = maximum factored moment at section due to externally applied loads (Live Loads)
(kip-in)
Mdnc = total unfactored dead load moment acting on the noncomposite section (kip-ft)
Values for the following moments and shears are at the critical section, Lcrit 6.25 feet from
the end of the girder at the abutment.
Vd 141 kips
Vi 136 kips
fr 0.20 f'c λ = 1.0 (normal wgt. conc.) LRFD [5.4.2.8] fr 0.566 ksi
T 1602 kips
T T es_crit
fcpe fcpe 3.548 ksi
Ag Sb
Snc Sb in3
Snc 18825
12Mdnc
Mcre Sc fr fcpe Mcre 89892 kip-in
Snc
Calculate Vci , LRFD [5.8.3.4.3] λ = 1.0 (normal wgt. conc.) LRFD [5.4.2.8]
Vci max Vci1 Vci2 Vci 1384.0 kips
T T es_crit Mdnc 12
ft ksi
Ag St St ft 0.340
T T es_crit Mdnc 12
fb fb 3.076 ksi
Ag Sb Sb
ycgb 48.76 in
ht 72.00 in
ft fb
fpc fb ycgb fpc 1.223 ksi
ht
slope
Vp_cw nsd As fpe Vp_cw 29.4 kips
100
Calculate Vcw , LRFD [5.8.3.4.3] λ = 1.0 (normal wgt. conc.) LRFD [5.4.2.8]
Vcw 0.06 λ f'c 0.30 fpc bv dv Vp_cw Vcw 256.1 kips
Vc min Vci Vcw Vc 256.1 kips
Vu_crit γstDC VDCnc VDCc γstDW VDWc γstLL VuLL where,
VDCnc 123.357 kips VDCc 8.675 kips VDWc 8.967 kips VuLL 100.502 kips
kips Vu_crit
Vu_crit 354.368 kips
Vn Vn 393.7
ϕv
fy 60 ksi
dv 65.00 in
fpc
min 1.0 3 1.8 otherwise
f' c
cotθ
Vs = Av fy dv LRFD Eq 5.8.3.3-4 reduced per
s C5.8.3.3-1 when = 90 degrees.
cotθ
s Av fy dv s 20.399 in
Vs
smax1
min 0.8 dv 18 if vu 0.125 f'c smax1 18.00 in
smax min smax1 smax2 smax 18.00 in
cotθ
Vs Av fy dv kips
s Vs 156
Check Vn requirements:
Vn min Vn1 Vn2 Vn 412 kips
Vu 0.5 ϕv Vc Vp (all values shown are in kips)
Use #4 U shaped stirrups at 18-inch spacing between the typical end sections. Unless a large
savings in rebar can be realized, use a single stirrup spacing between the standard end
sections.
Mmax Vu_crit
T ps 0.5 Vs Vp_cw cotθ T ps 670 kips
dv ϕf ϕv
Is the capacity of the straight strands greater than T ps? check "OK"
The strand is anchored lpx 10 inches. The transfer and development lengths for a
prestressing strand are calculated in accordance with LRFD [5.11.4.2]:
ltr 36.00 in
ld 146.2 in
Since lpx is less than the transfer length, the design stress in the prestressing strand is
calculated as follows:
The assumed crack plane crosses the centroid of the straight strands at
fpe lpx'
fpb fpb 78.37 ksi
60 db
Tendon capacity of the straight strands: nss As fpb 646 kips
The values of Vu , Vs, Vp and may be taken at the location of the critical section.
Over the length dv, the average spacing of the stirrups is:
6 4.25 6 5.5
save save 4.88 in
12
cotθ
Vs Av fy dv Vs 576 kips
save
The factored shear force at the critical section is: Vu_crit 354 kips
Vu_crit
T breqd 0.5 Vs Vp_cw cotθ T breqd 137 kips
ϕv
Is the capacity of the straight strands greater than T breqd ? check "OK"
dv 65.00 in
Vu_crit
vui vui 0.303 ksi
bvi dv
Vn = c Acv μ Avf fy Pc LRFD [5.8.4.1]
The nominal shear resistance, Vn , used in design shall not be greater than the lesser of:
c 0.28 ksi
μ 1.0
K1 0.3
K2 1.8
ts
wc
12
Pcd
2 S
S soh 2 Pcd 0.667 klf
hau wtf
Pch wc Pch 0.100 klf
2
12
From earlier calculations, the maximum #4 stirrup spacing used is s 18.0 inches.
Av
Avf 12 Avf 0.267 in2 /ft
s
Vn c Acv μ Avf fy Pc Vn 77.5 kips/ft
Vn min Vn Vn1 Vn2 Vn 77.5 kips/ft
Solution:
#4 stirrups spaced at s 18.0 inches is adequate to develop the required interface shear
L 12
Δlimit Δlimit 2.190 inches
800
Icg 1203475.476
ng 6 number of girders
From CBA analysis with 3 lanes loaded, the truck deflection controlled:
Δtruck 0.648 in
Applying the multiple presence factor from LRFD Table [3.6.1.1.2-1] for 3 lanes loaded:
Is the actual deflection less than the allowable limit, < limit? check "OK"
Pi_s nss As ftr ΔfpES Pi_s 1520 kips
Pi_d nsd As ftr ΔfpES Pi_d 319.971 kips
A 67.000 in
C 5.000 in
M2 Pi_d ( A C)
M2 19838.175 kip-in
M3 Pi_d A yb M3 10280.654 kip-in
The downward deflection due to the dead load of the deck and diaphragms:
Calculate the additional non-composite dead loads for an interior girder:
4
5 wnc L
3
Δnc 12 Δnc 2.161 in
384 EB Ig
The downward deflection due to the dead load of the parapets is calculated as follows. Note
that the deflections due to future wearing surface loads are not considered.
Calculate the composite dead loads for an interior girder:
wws 0 klf
4
5 wc L
3
Δc 12 Δc 0.173 in
384 EB Icg
The total downward deflection due to dead loads acting on an interior girder:
ΔDL Δnc Δc in
ΔDL 2.334
RC Δi ΔDL in
RC 1.005
Table of Contents
E19-2 Two-Span 54W" Girder, Continuity Reinforcement LRFD................................................................. 2
E19-2.1 Design Criteria ...................................................................................................... 2
E19-2.2 Modulus of Elasticity of Beam and Deck Material................................................. 3
E19-2.3 Section Properties ................................................................................................ 3
E19-2.4 Girder Layout ........................................................................................................ 4
E19-2.5 Loads .................................................................................................................... 4
E19-2.5.1 Dead Loads .......................................................................................... 4
E19-2.5.2 Live Loads ............................................................................................ 5
E19-2.6 Load Distribution to Girders .................................................................................. 5
E19-2.6.1 Distribution Factors for Interior Beams: ................................................ 6
E19-2.6.2 Distribution Factors for Exterior Beams: ............................................... 7
E19-2.7 Load Factors......................................................................................................... 9
E19-2.8 Dead Load Moments ............................................................................................ 9
E19-2.9 Live Load Moments ..............................................................................................10
E19-2.10 Factored Moments..............................................................................................10
E19-2.11 Composite Girder Section Properties .................................................................11
E19-2.12 Flexural Strength Capacity at Pier ......................................................................12
E19-2.13 Bar Cut Offs ........................................................................................................17
WisDOT Bridge Manual Chapter 19 – Prestressed Concrete
40'-0" Clear
130 ft 130 ft
CL Pier
Lg 130.375 total length of the girder (the girder extends 6 inches past the center
of bearing at the abutment and 1.5" short of the center line of the
pier).
ts 8 slab thickness, in
wc 0.150 kcf
f' c 1000
Ebeam8 5500 Ebeam8 6351
6000 EB Ebeam8
ED Edeck4
EB
n
ED n 1.540
tt 4.625 in
tw 6.5 in tw
tb
tb 10.81 in
ht 54 in
bw 30 width of bottom flange, in
Ag 798 in2
Ig 321049 in4
yt 27.70 in tse
eg yt 2 eg 33.45 in
yb 26.30 in 2
Sb 12205 in3
Kg n Ig Ag eg
2
Kg 1868972 in4
Chapter 19 suggests that at a 130 foot span, the girder spacing should be 7'-6" with 54W
girders.
S 7.5 ft
Assume a minimum overhang of 2.5 feet (2 ft flange + 6" overhang), soh 2.5
wb soh
ns ns 5.333
S
Use the lowest integer: ns floor ( ns) ns 5
Number of girders: ng ns 1 ng 6
wb S ns
Overhang Length: soh soh 2.50 ft
2
E19-2.5 Loads
wdlxi wg wd
S s w 2 wdx
oh h wdlxi 1.559 klf
2 L
interior:
wdi
wdlii wg wd S wh 2 wdlii 1.687 klf
L
2 wp
wp wp 0.129 klf
ng
w wws
wws wws 0.133 klf
ng
* LRFD [4.6.2.2.1] states that permanent loads on the deck may be distributed uniformly
among the beams. This method is used for the parapet and future wearing surface loads.
DLA of 33% applied to truck or tandem, but not to lane per LRFD [3.6.2.1].
For Fatigue 1:
HL-93 truck (no lane) with 15% DLA and 30 ft rear axle spacing per LRFD [3.6.1.4.1].
Distribution factors are in accordance with LRFD [Table 4.6.2.2.2b-1]. For an interior beam,
the distribution factors are shown below:
Criteria for using distribution factors - Range of Applicability per LRFD [Table 4.6.2.2.2b-1].
NoBeams "OK" if ng 4
"NG" otherwise
S DeckSpan
7.5 "OK"
tse DeckThickness 7.5 "OK"
x L BridgeSpan x 130.0 "OK"
ng NoBeams 6.0 "OK"
Kg LongitStiffness
1868972.4 "OK"
Note: The distribution factors above already have a multiple lane factor included that is used
for service and strength limit states. The distribution factor for One Lane Loaded should be
used for the fatigue vehicle and the 1.2 multiple presence factor should be divided out.
wb w
wparapet Width of parapet wparapet 1.250 ft
2 overlapping the deck
Per LRFD [Table 4.6.2.2.2d-1] the distribution factor shall be calculated by the Lever Rule.
Calculate the distribution factor by the Lever Rule:
sw1 sw2
2'-0" 6'-0"
de S
S sw1 sw2
Rx Rx 0.500 % of a lane load
S 2
Add the single lane multi-presence factor, m 1.2
The exterior girder distribution factor is the maximum value of the One Lane Loaded case and
the Two or More Lanes Loaded case:
Note: The interior girder has a larger live load distribution factor and a larger dead load than
the exterior girder. Therefore, for this example, the interior girder is likely to control.
Tenth DC DC DW
Point non-composite composite composite
0.5 3548 137 141
0.6 3402 99 102
0.7 2970 39 40
0.8 2254 -43 -45
0.9 1253 -147 -151
1.0 0 -272 -281
The DCnc values are the component non-composite dead loads and include the weight of the
girder, haunch, diaphragms and the deck.
The DCc values are the component composite dead loads and include the weight of the
parapets.
The DWc values are the composite dead loads from the future wearing surface.
Note that the girder dead load moments (a portion of DCnc) are calculated based on the CL
bearing to CL bearing length. The other DCnc moments are calculated based on the span
length (center of bearing at the abutment to centerline of the pier).
The unfactored live load moments (per lane including impact) are listed below (values are in
kip-ft). Note that the impact factor is applied only to the truck portion of the HL-93 loads. A
separate analysis run will be required if results without impact are desired.
The unfactored live load moments per lane are calculated by applying the appropriate
distribution factor to the controlling moment. For the interior girder:
gi 0.619
The single lane distribution factor should be used and the multiple presence factor of 1.2 must
be removed from the fatigue moments.
1
MLLfatigue gi1 952.64 kip-ft
1.2 MLLfatigue 339
1.0 1.0 MDCc 1.0 MDWc 1.0 MLL Ms1 2608 kip-ft
Fatigue 1
we S 12 we 90.00 in
The effective width, we , must be adjusted by the modular ratio, n 1.54 , to convert to the
same concrete material (modulus) as the girder.
we
weadj weadj 58.46 in
n
hc 63.50 in
n 1.540
Note: The area of the concrete haunch is not included in the calculation of the composite
section properties.
ΣA 1236 in2
ΣAY
ycgb ycgb 38.2 in
ΣA
Acg ΣA in2
2
Icg ΣIplusAYsq Acg ycgb Icg 639053 in4
Deck:
Icg
Sc n in4
ycgt hau tse Sc 38851
BarNo 9
BarD BarNo
de ht hau ts cover BarD bartrans in
2 de 60.31
Mu 12
Ru Ru 0.532 ksi
2
ϕf bw de
f' c 2 Ru
ρ 0.85 1 1 ρ 0.00925
fy 0.85 f' c
This reinforcement is distributed over the effective flange width calculated earlier,
we 90.00 inches. The required continuity reinforcement in in2 /ft is equal to:
As
Asreq Asreq 2.232 in2 /ft
we
12
From Chapter 17, Table 17.5-3, for a girder spacing of S 7.5 feet and a deck thickness of
ts 8.0 inches, use a longitudinal bar spacing of #4 bars at slongit 8.5 inches. The
continuity reinforcement shall be placed at 1/2 of this bar spacing, .
#9 bars at 4.25 inch spacing provides an Asprov 2.82 in2 /ft, or the total area of steel
provided:
we
As Asprov As 21.18 in2
12
This is within the thickness of the bottom flange height of 7.5 inches.
c
If < 0.6 for (fy = 60 ksi) LRFD [5.7.2.1], the reinforcement has yielded and the
d
s assumption is correct.
a
LRFD [5.7.2.2] β 0.65 ; c
1 β c 9.582 in
1
c
= 0.16 < 0.6 therefore, the reinforcement will yield
d
s
Mn As fy de
a 1
Mn 6056 kip-ft
2 12
Mr ϕf Mn Mr 5451 kip-ft
Mu 4358 kip-ft
| fr 0.24 f' cd λ = 1.0 (normal wgt. conc.) LRFD [5.4.2.8] fr 0.480 ksi
Where:
γ1 1.6 flexural cracking variability factor
1
Mcr 1.1fr Sc Mcr 1709 kip-ft
12
Is Mr greater than the lesser value of Mcr and 1.33*Mu ? check "OK"
Check the Service I crack control requirements in accordance with LRFD [5.7.3.4]:
As
ρ ρ 0.01170
bw de
Es
n
EB n 4.566
2
k ( ρ n ) 2 ρ n ρ n k 0.278
k
j 1 j 0.907
3
Note that the value of dc should not include the 1/2-inch wearing surface.
BarD BarNo
dc cover 0.5 BarD bartrans dc 3.19 in
2
Ms1
fs 12 < 0.6 fy fs 27.006 ksi < 0.6 fy O.K.
As j de
The height of the composite section, h, is:
dc
β 1 β 1.076
0.7 h dc
700γe
Smax 2 dc Smax 11.70 in
β fs
spa 4.25 in
fmin is equal to the stress in the reinforcement due to the moments from the permanent loads
combined with the Fatigue I load combination. f is the stress range resulting from the fatigue
vehicle.
Check stress in section for determination of use of cracked or uncracked section properties:
Mf
ftop 12 ftop 0.328 ksi
Sc
If we assume the neutral axis is in the bottom flange, the distance from cracked section neutral
axis to bottom of compression flange, ycr, is calculated as follows:
2
bw ycr
= n As de ycr
2
n As 2 bw de
ycr 1 1 ycr 16.756 in No Good
bw n As
tbf_min 7.5
2
tbf_min x
wtaper tbf_min x 2 tw 2 =0
wtaper ttaper ttaper
x tbf_min n As de x
2 3
3 2 3
wtaper tbf_min tbf_min tweb x
Icr wtaper tbf_min x
12 2 3
3 2
wtaper ttaper wtaper ttaper ttaper
n As de x
2
x tbf_min
36 2 2
Distance from centroid of tension reinforcement to the cracked section neutral axis:
Mf yrb
fmin n 12 fmin 10.913 ksi
Icr
MLLfatigue yrb
Δf n 12 Δf 3.488 ksi
Icr
γfLL Δf 5.232 ksi
As
As' As' 10.588 in2
2
As' fy
a' a' 3.11 in
α bw f'c
1
0
(kip-ft)
3
2 10
3
4 10
3
6 10
0.7 0.8 0.9 1
Based on the moment diagram, try locating the first cut off at cut1 0.90 span. Note that the
Service I crack control requirements control the location of the cut off.
Is Mr' greater than the lesser value of Mcr and 1.33*Mucut1 ? check "OK"
Check the Service I crack control requirements in accordance with LRFD [5.7.3.4]:
As'
ρ' ρ' 0.00585
bw de
2
k' ( ρ' n) 2 ρ' n ρ' n k' 0.206
k'
j' 1 j' 0.931
3
Mscut1
fs' 12 < 0.6 fy fs' 31.582 ksi < 0.6 fy O.K.
As' j' de
β 1.076
γe 0.750
700γe
Smax' 2 dc Smax' 9.08 in
β fs'
spa' 8.50 in
Check stress in section for determination of use of cracked or uncracked section properties:
Mf cut1
ftop_cut1 12 ftop_cut1 0.234 ksi
Sc
The live load range is the sum of the positive and negative fatigue moments:
Mf LLrange
γfLLΔf_cut1 n 12 γfLLΔf_cut1 0.844 ksi
Sc
Therefore this cut off location, cut1 0.90 , is OK. The bar shall be extended past the cut off
point a distance not less than the maximum of the following, LRFD [5.11.1.2.3]:
de 60.311
extend 12 BarD BarNo extend 13.536
0.0625 L 12 97.500
max ( extend)
8.13 ft
12
max ( extend)
X1 L 1 cut1 X1 21.12 feet
12
USE X1 22 feet from the CL of the pier.
The second bar cut off is located at the point of inflection under a uniform 1.0 klf composite
dead load. At cut2 0.750 , Mcut2 ( 79 ) kip-ft. Extend the bar the max(extend) distance
calculated above past this point, or 4 feet past the first cut off, whichever is greater.
(kip-ft) 0
3
1 10
3
2 10
3
3 10
0.7 0.8 0.9 1
max ( extend)
X2a L 1 cut2 X2a 40.63 feet from the center of the pier
12
Table of Contents
E19-3 Box Section Beam ............................................................................................................................. 2
E19-3.1 Preliminary Structure Data.................................................................................... 2
E19-3.2 Live Load Distribution ........................................................................................... 4
E19-3.2.1 Distribution for Moment......................................................................... 4
E19-3.2.2 Distribution for Shear ............................................................................ 5
E19-3.3 Live Load Moments .............................................................................................. 7
E19-3.4 Dead Loads .......................................................................................................... 7
E19-3.5 Dead Load Moments ............................................................................................ 9
E19-3.6 Design Moments...................................................................................................10
E19-3.7 Load Factors.........................................................................................................11
E19-3.8 Factored Moments................................................................................................11
E19-3.9 Allowable Stress ...................................................................................................12
E19-3.9.1 Temporary Allowable Stresses .............................................................12
E19-3.9.2 Final Condition Allowable Stresses ......................................................12
E19-3.10 Preliminary Design Steps ...................................................................................13
E19-3.10.1 Determine Amount of Prestress..........................................................13
E19-3.10.1.1 Estimate the Prestress Losses ...........................................14
E19-3.10.1.2 Determine Number of Strands ............................................15
E19-3.10.2 Prestress Loss Calculations ...............................................................16
E19-3.10.2.1 Elastic Shortening Loss ......................................................16
E19-3.10.2.2 Approximate Estimate of Time Dependant Losses.............17
E19-3.10.3 Check Stresses at Critical Locations ..................................................18
E19-3.11 Flexural Capacity at Midspan .............................................................................21
E19-3.12 Shear Analysis....................................................................................................23
E19-3.13 Non-Prestressed Reinforcement (Required near top of girder) ........................29
E19-3.14 Longitudinal Tension Flange Capacity:...............................................................30
E19-3.15 Live Load Deflection Calculations.......................................................................31
E19-3.16 Camber Calculations ..........................................................................................32
WisDOT Bridge Manual Chapter 19 – Prestressed Concrete
fpu 270 Ultimate tensile strength for low relaxation strands, ksi
wc 0.150 Unit weight of concrete for box girder, overlay, and grout, kcf
Based on past experience, the modulus of elasticity for the precast concrete are given in
Chapter 19 as Ebeam6 5500 ksi for a concrete strength of 6 ksi. The values of E for
different concrete strengths are calculated as follows:
f' c 1000
Ebeam5 5500 Ebeam5 5021 ksi
6000
EB Ebeam5
The modulus of elasticity at the time of release is calculated in accordance with LRFD
[C5.4.2.4].
1.5
Ebeam4.25 33000 K1 wc f'ci Ebeam4.25 3952 ksi
Ect Ebeam4.25
Based on the preliminary data, Section 19.3.9 of this chapter and Table 19.3-3, select a 4'-0"
wide pretensioned box section having a depth of 1'-9" (Section 3), as shown on Bridge Manual
Standard 19.15. The actual total deck width provided is calculated below.
nbeams 8
Ws 4 Width of section, ft
Wcurb 1.5 Width of curb on exterior girder (for steel rails), feet
Wj
S Ws Effective spacing of sections S 4.125 feet
12
yt 10.5 in
yb 10.5 in
( 1 μ) I
K K 0.759
J
Wb
C min K K C 0.567
L
When C is less than 5:
2
D 11.5 NL 1.4 NL ( 1 0.2 C) D 11.701
S
gint_m gint_m 0.353
D
For exterior beams, the live load moment distribution factor is calculated as indicated in LRFD
[Table 4.6.2.2.2d-1] for cross section type "g".
5 1
de Whrail
12 2
de
de 0.212 feet
de
e1 max 1.125 1 e1 1.118
30
de
e2 max 1.04 1 e2 1.032
25
Use the maximum value from the above calculations to determine the controlling exterior girder
distribution factor for moment.
gext_m max gext1 gext2 gext_m 0.394
The distribution factor for fatigue is the single lane distribution factor with the multi-presence
factor, m 1.2, removed:
gext1
gf gf 0.328
1.2
For the single lane loaded, only one wheel can be located on the box section. With the
single lane multi presence factor, the interior girder shear distribution factor is:
For two or more lanes loaded, center adjacent vehicles over the beam. One load from
each vehicle acts on the beam.
4'-0"
2'-1 ½” 2'-1 ½”
4'-1 ½”
Equivalent Girders for Simple Span
Distribution
2.125
gint_v2 0.5 2 gint_v2 0.515
4.125
gint_v max gint_v1 gint_v2 gint_v 0.600
Exterior Girder
For the exterior girder, the range of applicability of LRFD [T-4.6.2.2.3b-1] for bridge type
"g" is satisfied.
For the single lane loaded:
de
ev1 max 1.25 1.0 ev1 1.239
20
gext_v1 ev1 gint_v1
gext_v1 0.744
0.5
2.0
b
de
12
ev2 max 1 1.0 ev2 1.211
40
gext_v max gext_v1 gext_v2 gext_v 0.744
The HL-93 live load moment per lane on a 44 foot span is controlled by the design tandem
plus lane. The maximum value at mid-span, including a dynamic load allowance of 33%, is
MLL_lane 835.84 kip-ft per lane.
The Fatige live load moment per lane on a 44 foot span at mid-span, including a dynamic load
allowance of 15%, is MLLfat_lane 442.4 kip-ft per lane.
For the interior girders, all dead loads applied after the post tensioning has been completed
are distributed equally to all of the girders.
toverlay
Overlay
12
Wb Wcurb 2 wc
wo wo 0.093 klf
nbeams
Wj toverlay njoints
Joint Grout wj Ds wc wj 0.031 klf
12 12 nbeams
2 wrail
"M" Rail wr wr 0.019 klf
nbeams
Future Wearing Surface
Wb 0.020
wfws wfws 0.082 klf
nbeams
For the exterior girders, all dead loads applied directly to the girder are applied.
toverlay
Overlay wo_ext
12
S Wcurb wc wo_ext 0.066 klf
1 Wj toverlay
Joint Grout wj_ext Ds w wj_ext 0.018 klf
2 12 12 c
Type M Rail wr_ext wrail wr_ext 0.075 klf
2
L
Interior Girder MDCint wDCint MDCint 191.8 kip-ft
8
2
L
MDWint wDWint MDWint 19.9 kip-ft
8
2
L
Exterior Girder MDCext wDCext MDCext 204.5 kip-ft
8
2
L
MDWext wDWext MDWext 20.0 kip-ft
8
Since the Dead Load moments are very close and the exterior Live Load moments are greater
than the interior moments, the exterior girder controls for this design example. Note: an interior
box girder section design will not be provided in this example. However, the interior girder shall
not have less load carrying capacity then the exterior girder.
Strength 1
Mstr η γstDC MDC γstDW MDW γstLL MLL
1.0 1.25 MDC 1.50 MDW 1.75 MLL Mstr 862 kip-ft
Ms1 η γs1DC MDC γs1DW MDW γs1LL MLL
1.0 1.0 MDC 1.0 MDW 1.0 MLL Ms1 554 kip-ft
Ms3 η γs3DC MDC γs3DW MDW γs3LL MLL
1.0 1.0 MDC 1.0 MDW 0.8 MLL Ms3 488 kip-ft
In accordance with LRFD [Table 5.9.4.1.2-1], the temporary allowable tension stress is
calculated as follows (assume there is no bonded reinforcement):
ftiall min 0.0948 λ f' ci 0.2 λ = 1.0 (normal wgt. conc.)
LRFD [5.4.2.8]
ftiall 0.195 ksi
If bonded reinforcement is present in the top flange, the temporary allowable tension stress is
calculated as follows:
ftiall_bond 0.24 λ f' ci λ = 1.0 (normal wgt. conc.) ftiall_bond 0.495 ksi
LRFD [5.4.2.8]
ftall 0.19 f' c LL + PS + DL | ftall | < 0.6 ksi ftall 0.425 ksi
1) Design the amount of prestress to prevent tension at the bottom of the beam under the full
| load at center span after losses.
2) Calculate the prestress losses and check the girder stresses at mid span at the time of
transfer.
3) Check resulting stresses at the critical sections of the girder at the time of transfer (before
| losses and while in service (after losses).
Near center span, after losses, T = the remaining effective prestress, aim for no tension at the
| bottom. Use Service I for compression and Service III for tension.
For this example, the exterior girder has the controlling moments.
Calculate the stress at the bottom of the beam due to the Service 3 loading:
Ms3 12
fb fb 1.867 ksi
Sb
Stress at bottom due to prestressing:
T yb
fbp = 1 e
A 2
r
and fbp fb desired final prestress.
We want this to balance out the tensile stress calculated above from the loading, i.e. an initial
compression. The required stress due to prestress force at bottom of section to counteract the
Service 3 loads:
fbp 1.867 ksi
Note: The estimated time dependant losses (based on experience for low
relaxation strands) will be re-calculated using the approximate method in
accordance with LRFD [5.9.5.3] once the number of strands has been determined.
Assume an initial strand stress; ftr 0.75 fpu ftr 202.5 ksi
Based on experience, assume ΔfpES_est 9.1 ksi loss from elastic shortening. As an
alternate initial estimate, LRFD [C.5.9.5.2.3a] suggests assuming a 10% ES loss.
ΔfpES_est
ESloss 100 ESloss 4.494 %
ftr
The total loss is the time dependant losses plus the ES losses:
loss
loss% 100 loss% 19.309 % (estimated)
ftr
T = 1 loss% To
loss%
ratio 1 ratio 0.807
100
T = ratio To
1 loss% To 1 e yb
fbp =
A 2
r
OR:
fbp To yb
= 1 e
1 loss% A 2
r
fbp
fbpi
loss% fbpi 2.313 ksi
1
100
desired bottom initial prestress
As 0.153 in2
P N yb
fbp 1 e
A rsq
yb 10.5 Distance from the centroid of the 21" depth to the bottom of the
box section, in.
For the 4'-0 wide box sections, there can be up to 22 strands in the bottom row and 2 rows
of strands in the sides of the box. Calculate the eccentricity for the maximum number of
strands that can be placed in the bottom row of the box:
es eb inches
es 8.5
fbpi A 1
Nreq
P yb Nreq 17 strands
1 es
rsq
Therefore, try N 16 strands
since some final tension in the
bottom of the girder is allowed.
eb 14 eb 2 2
es
16
es 8.25 inches
1) Elastic Shortening (ES), shortening of the beam as soon as prestress is applied. Can this
be compensated for by overstressing?
3) Creep (CR), slow shortening of concrete due to permanent compression stresses in the
beam, time function.
4) Relaxation (RE), the tendon slowly accommodates itself to the stretch and the internal
stress drops with time
The ES loss estimated above was: ΔfpES_est 9.1 ksi, or ESloss 4.494 %. The resulting
force in the strands after ES loss:
ESloss
T o 1 Toi T o 474 kips
100
Since all strands are straight, we can calculate the initial elastic shortening loss;
To es es
fcgp
A
T o es
I
Mgi 12
I
fcgp 1.264 ksi
Ep Es Ep 28500 ksi
Ep
ΔfpES f ΔfpES 9.118 ksi
Ect cgp
This value of fpES is in agreement with the estimated value above; ΔfpES_est 9.10 ksi. If
these values did not agree, T o would have to be recalculated using ftr minus the new value of
fpES, and a new value of fcgp would be determined. This iteration would continue until the
assumed and calculated values of fpES are in agreement.
Check the design to avoid premature failure at the center of the span at the time of transfer.
Check the stress at the center span (at the plant) at both the top and bottom of the girder.
To T o es Mgi 12
fttr fttr 0.200 ksi
A St St
To T o es Mgi 12
fbtr fbtr 1.392 ksi
A Sb Sb
Is the stress at the top of the girder less than the allowable? check "OK"
Is the stress at the bottom of the girder less than the allowable? check "OK"
fpi As
ΔfpLT = 10.0 γh γst 12.0 γh γst Δf
Ag pR
From LRFD [Figure 5.4.2.3.3-1], the average annual ambient relative humidity, H 72 %.
5
γst γst 0.952
1 f' ci
ftr As N
ΔfpCR 10.0 γh γst ΔfpCR 7.781 ksi
A
Δfp
100 15.061 % total
ftr prestress loss
This value is less than but in general agreement with the initial estimated loss% 19.3 .
The remaining stress in the strands and total force in the beam after all losses is:
ltr
x x 2.50 feet
12
The moment at the end of the transfer length due to the girder dead load:
wg_ext wdiaph x
Lg x x
2
Mgt
2
2
wdiaph_end x
Mgt 38 kip-ft
To T o es Mgt 12
ftt ftt 0.303 ksi
A St St
ftiall 0.195 ksi
check "NG"
If bonded reinforcement is provided in the top flange, the allowable stress is:
ftiall_bond 0.495 ksi
To T o es Mgt 12
fbt fbt 1.896 ksi
A Sb Sb
fciall 2.763 ksi
Check final stresses after all losses at the mid-span of the girder:
T T es Ms1 12
ft2 LL + PS + DL ft2 1.719 ksi
A St St
check "OK"
T T es
MDC MDW 12
ksi
fb1 PS + DL
fb1 0.958
A Sb Sb
check "OK"
T T es Ms3 12
fb fb 0.051 ksi
A Sb Sb
check "OK"
1 T T es
MDC MDW 12 MLLf 12
ftf1 1/2(PS + DL) + LLf
2 A St St St
check "OK"
fps = fpu 1 k
c
dp
where:
fpy
k = 2 1.04
fpu
"c" is defined as the distance between the neutral axis and the compression face (inches).
Assume that the compression block is in the top section of the box. Calculate the capacity as if
it is a rectangular section. The neutral axis location,calculated in accordance with LRFD
5.7.3.1.1 for a rectangular section, is:
Aps fpu
c=
fpu
α1 f'c β1 b k Aps
dp
where:
Aps N As Aps 2.45 in2
b Ws 12 b 48.00 in
β1 max 0.85 f' c 4 0.05 0.65
β1 0.800
dp yt es dp 18.75 in
Aps fpu
c
fpu c 3.82 in
α1 f' c β1 b k Aps
dp
a β1 c a 3.06 in
This is within the depth of the top slab (5-inches). Therefore our assumption is OK.
c
fps fpu 1 k
dp fps 254.6 ksi
T u fps Aps T u 624 kips
Calculate the nominal moment capacity of the section in accordance with LRFD [5.7.3.2]:
a 1
Mn Aps fps dp
2 12
Mn 895 kip-ft
For prestressed concrete, ϕf 1.00, LRFD [5.5.4.2.1]. Therefore the usable capacity is:
Check the section for minimum reinforcement in accordance with LRFD [5.7.3.3.2] for the
interior girder:
1.33 Mu 1147 kip-ft
fr 0.24 f'c λ = 1.0 (normal wgt. conc.) LRFD [5.4.2.8] fr 0.537 ksi
T T es
fcpe fcpe 1.816 ksi
A Sb
Is Mr greater than the lesser value of Mcr and 1.33*Mu ? check "OK"
The live load shear distribution factors to the girders are calculated above in E19-3.2.2.
gint_v 0.600
gext_v 0.744
From section E19-3.4, the uniform dead loads on the girders are:
Interior Girder wDCint 0.792 klf
However, the internal concrete diaphragms were applied as total equivalent uniform loads to
determine the maximum mid-span moment. The diaphragm weights should be applied as poin
loads for the shear calculations.
100
(kips)
Vu
50
0
0 0.1 0.2 0.3 0.4 0.5
TenthPoints Vu0.0 143.5 kips
bv 2tw bv 10.00 in
The critical section for shear is taken at a distance of dv from the face of the support, LRFD
[5.8.3.2].
dv = effective shear depth taken as the distance between the resultants of the tensile and
compressive forces due to flexure. It need not be taken less than the greater of 0.9*de or
0.72h (inches). LRFD [5.8.2.9]
a
dv es yt dv 17.22 in
2
For the standard bearing pad of width, wbrg 8 inches, the distance from the end of the
girder to the critical section:
1
Lcrit wbrg dv
12 Lcrit 2.10 ft
Also, the value of fpu , should be revised if the critical section is located less than the
development length from the end of the beam. The development length for a prestressing
strand is calculated in accordance with LRFD [5.11.4.2]:
ld K fps fpe ds
2
ld 70.0 in
3
The transfer length may be taken as: ltr 60 ds ltr 30.00 in
Since Lcrit 2.102 feet is between the transfer length and the development length, the
design stress in the prestressing strand is calculated as follows:
Lcrit 12
fpu_crit fpe fpu_crit 145 ksi
ltr
dv dv_crit
The location of the critical section from the end of the girder is:
1
Lcrit wbrg dv
12
Lcrit 2.159 ft
The location of the critical section from the center line of bearing at the abutment is:
crit Lcrit 0.25 crit 1.909 ft
The nominal shear resistance of the section is calculated as follows, LRFD [5.8.3.3]:
Vn = min Vc Vs Vp 0.25 f'c bv dv Vp
Note, the value of Vp does not equal zero in the calculation of Vcw.
Vd = shear force at section due to unfactored dead load and includes both DC and DW (kips)
Vi = factored shear force at section due to externally applied loads (Live Loads) occurring
simultaneously with Mmax (kips). (Not necessarily equal to Vu .)
Mcre = moment causing flexural cracking at section due to externally applied loads (kip-in)
Mmax = maximum factored moment at section due to externally applied loads (Live Loads)
(kip-in)
Mdnc = total unfactored dead load moment acting on the noncomposite section (kip-ft)
Values for the following moments and shears are at the critical section, Lcrit 2.16 feet from
the end of the girder at the abutment.
Vd 18.3 kips
Vi 109.5 kips
fr 0.20 f'c λ = 1.0 (normal wgt. conc.) LRFD [5.4.2.8] fr 0.447 ksi
T 421 kips
T crit T crit es
fcpe fcpe 1.527 ksi
A Sb
Mdnc 37 kip-ft
Sc Sb Sc 3137 in3
12Mdnc
Mcre Sc fr fcpe Mcre 5746 kip-in
Snc
Calculate Vci , LRFD [5.8.3.4.3] λ = 1.0 (normal wgt. conc.) LRFD [5.4.2.8]
Vi Mcre
Vci2 0.02 λ f'c bv dv Vd Vci2 495.9 kips
Mmax
Vci max Vci1 Vci2 Vci 495.9 kips
yb 10.50 in
h 21.00 in
ft fb
fpc fb yb fpc 0.595 ksi
h
Calculate Vcw , LRFD [5.8.3.4.3] λ = 1.0 (normal wgt. conc.) LRFD [5.4.2.8]
Vcw 0.06 λ f'c 0.30 fpc bv dv Vp_cw Vcw 56.0 kips
Vc min Vci Vcw Vc 56.0 kips
Vu_crit
Vn Vn 147.6 kips
ϕv
fy 60 ksi
dv 17.91 in
fpc
min 1.0 3 1.8 otherwise
f' c
cotθ
Vs = Av fy dv LRFD Eq 5.8.3.3-4 reduced per
s C5.8.3.3-1 when = 90 degrees.
cotθ
s Av fy dv s 8.441 in
Vs
smax1
min 0.8 dv 24 if vu 0.125 f'c smax1 7.16 in
smax min smax1 smax2 smax 7.16 in
cotθ
Vs Av fy dv kips
s Vs 110.4
Check Vn requirements:
Vn min Vn1 Vn2 Vn 166 kips
Vu 0.5 ϕv Vc Vp (all values shown are in kips)
Use #4 U shaped stirrups at 7-inch spacing between the typical end sections. Unless a large
savings in rebar can be realized, use a single stirrup spacing between the standard end
sections.
Ds 1.75 feet
Ds
b 48 inches
ftb
ftt Ds 12
Y Y 2.898 inches
ftt fbt
ftt b Y
T T 21.101 kips
2
fy 60
T
Areqd Areqd 0.703 in2
0.5 fy
Mmax Vu_crit
T ps Vp_cw 0.5 Vs cotθ T ps 241 kips
dv ϕf ϕv
Is the capacity of the straight strands greater than T ps? check "OK"
The strand is anchored lpx 8 inches. The transfer and development lengths for a
prestressing strand are calculated in accordance with LRFD [5.11.4.2]:
ltr 30.00 in
ld 70.0 in
Since lpx is less than the transfer length, the design stress in the prestressing strand is
calculated as follows:
The assumed crack plane crosses the centroid of the straight strands at
Ys yb es Ys 2.25 in
fpe lpx'
fpb fpb 69.07 ksi
60 ds
The values of Vu , Vs, Vp and may be taken at the location of the critical section.
Over the length dv, the average spacing of the stirrups is:
cotθ
Vs Av fy dv Vs 110 kips
save
The factored shear force at the critical section is: Vu_crit 133 kips
Vu_crit
T breqd 0.5 Vs Vp_cw cotθ T breqd 166 kips
ϕv
Is the capacity of the straight strands greater than T breqd ? check "OK"
L 12
Δlimit Δlimit 0.660 inches
800
I 32942 in4
nbeams 8
From CBA analysis with 2 lanes loaded, the truck deflection controlled:
Δtruck 0.347 in
Applying the multiple presence factor from LRFD [Table 3.6.1.1.2-1] for 2 lanes loaded:
Is the actual deflection less than the allowable limit, < limit? check "OK"
Pi_s N As ftr ΔfpES Pi_s 474 kips
2
M 1 Lg
2
Δs 12 Δs 1.07 in
8 Ect I
5 wg wd Lg
3 4
in
Δgi 12 Δgi 0.44
384 Ect I
Anticipated prestress camber at release:
Δi ΔPS Δgi Δi 0.63 in
The downward deflection due to the dead load of the joint grout, overlay, railing and future
wearing surface:
Calculate the additional non-composite dead loads for an exterior girder:
4
5 wnc L 3
Δnc 12 Δnc 0.123 in
384 EB I
Table of Contents
E19-4 Lifting Check for Prestressed Girders, LRFD ......................................................................................2
E19-4.1 Design Criteria ........................................................................................................2
E19-4.2 Lifting Stresses .......................................................................................................2
E19-4.3 Check Compression Stresses due to Lifting...........................................................4
E19-4.4 Check Tension Stresses due to Lifting ...................................................................4
E19-4.5 Design Top Flange Reinforcement .........................................................................4
WisDOT Bridge Manual Chapter 19 – Prestressed Concrete
girder_size "72W-inch"
tw 6.5 inches
1
R wgirder Lgirder R 69.569 kips
2
V1 R V1 69.569 kips
R R kips
V2 R V2 69.569
V1 V2
wgirder 0.1 Lgirder
Mgird0.1 Lgirder 0.1Lgirder
2
M0.5
M0.1
Mgird0.1 914.14 kip-ft
Mgird0.1 12
ftop_prestr fi_top_0.1 ftop_prestr 0.336 ksi
Stop
Mgird0.1 12
fbot_prestr fi_bot_0.1 fbot_prestr 4.062 ksi
Sbot
The girder dead load moment and resulting stresses are calculated based on the girder being
supported at the lift points. The resulting stresses are added to the stresses due to the
prestress forces to give the total stresses during girder picks.
Moments and Shears at the Lift Points, 1/10 point, due to the girder self weight.
R 69.569 kips
M0.1 kips
V'4 V'3 R V'4 13.914
1
Mgird0.1_Lift V'1 Lgirder 0.1 Mgird0.1_Lift 101.57 kip-ft
2
Top of girder stresses due to lifting forces (postive stress values indicate compression.):
Mgird0.1_Lift 12
ftop_Lift ftop_prestr ftop_Lift 0.405 ksi
Stop
Mgird0.1_Lift 12
fbot_Lift fbot_prestr fbot_Lift 4.126 ksi
Sbot
| fciall 0.65 f' ci where f'ci = 6.8 ksi fciall 4.42 ksi
Is the stress at the bottom of the girder less than the allowable? check f_bot "OK"
| If stress at the bottom of girder is greater than allowable, calculate f' ci_reqd
:
fbot_Lift
| f' ci_reqd
0.65
| ftall 0.24 λ f'ci λ = 1.0 (normal wgt. conc.) ftall 0.626 ksi
LRFD [5.4.2.8]
ftop_Lift 0.405 ksi
Is the stress at the top of the girder less than the allowable? check f_top "OK"
Therefore, proportion the reinforcement in the top flange using an allowable stress of:
hgirder 72 in ftop_Lift
y
fbot_Lift 4.126 ksi
hgirder
ftop_Lift
y hgirder
ftop_Lift fbot_Lift
y 6.441 in
t2 ttop_flg_max t2 5.5 in
A1 t1 A2 t2
ttop_flg_avg ttop_flg_avg 4.461 in
A1 A2
ttop_flg_avg
ftop_Lift
y fflg_avg
ftop_Lift
fflg_avg y ttop_flg_avg fflg_avg 0.125 ksi
y
wtop_flange/2
ftop_Lift
ttop_flg_avg
y
fflg_avg
1
T flg_avg ftop_Lift fflg_avg ttop_flg_avg wtop_flg T flg_avg 56.742 kips
2
Calculate the tension force in the girder web (this minor force can be ignored for simplification).
ftop_Lift
ttop_flg_avg
y
fflg_avg
tw/2
1
T web fflg_avg y ttop_flg_avg tw T web 0.802 kips
2
T 57.544 kips
T
AsReqd AsReqd 1.918 in2
fs
AsReqd
As As 0.32 in2 per bar
Number_Bars
Is the area of steel per bar greater than required? check As "OK"
| Therefore, use 6 - #6 Bars in Top Flange of Girder for 0.1 point lifting locations.
Note that these bars should be terminated where no longer required by design and lapped
with 6 #4 bars as shown on the Standard Details.